The Trump administration and its allies in Congress have conspired to slow the transition to clean energy in this country, undercutting solar and wind power, propping up coal plants, and backing fossil fuels over renewable energy at every turn. Clean energy, though, has pushed back, and pushed forward. Solar in particular came into the second Trump administration with a whole lot of momentum, and that momentum has carried renewable energy to many new heights lately.

Here are four examples of impressive renewable energy progress, how they came about, and what to watch for next.

US solar blows past its generation record

What happened: US solar generation in 2025 was a stunning 28% higher than in 2024. The electricity flowing from all our solar—on rooftops and parking lots, in fields and deserts—was equivalent to the amount used by every household across 14 states in the Midwest and Northeast, from Wisconsin to New Jersey and up to Maine. Solar generation in a single month (July 2025) was more than a full year’s worth of solar generation just a decade ago.

How it happened: A lot of solar capacity has come online lately. 2024 was a record year for new solar installations in the United States, 21% higher than the previous record-setter. While new installations in 2025 were down from that height, last year still had the second-highest tally ever. Texas installed a Texas-sized amount of that recent solar, California followed, and Indiana took the #3 spot, up from #15 in 2023. The total capacity by the end of 2025 was enough to meet the electricity needs of tens of millions of US households.

What’s next: The pace of new installations will be driven in part by deadlines and restrictions imposed by last year’s megabill, and will also depend on the vagaries of the tariff regime under this administration. This year is sure to break the record for solar generation yet again, though, with all of 2025’s installations leaping into action and more solar coming online every day. Preliminary results suggest that solar electricity generation year-to-date is more than 20% higher than in the same period last year.

US wind breaks record after record

What happened: Wind power is the largest source of renewable energy in this country, and it keeps growing. For example, last month in New England, wind farms added up to a record amount of peak generation—more than 30% above where the region’s wind record had stood just six months earlier. Plus, a record-breaking project is starting to come online in the US Southwest: SunZia will be the largest wind project in North America—and, with its associated transmission line, the largest US renewable energy project ever.

How it happened: InNew England’s case, a lot of credit goes to offshore wind. While the region’s land-based wind farms play an important role, what’s new are the injections of power from offshore wind turbines. Vineyard Wind, located in the waters off Massachusetts, has been generating power from a majority of its turbines since last year, and will be fully online very shortly. Once it’s completed, it will be the first large-scale offshore wind project in the United States—indeed, in all of the Americas. Add in strong offshore winds, and you have a recipe for some awesome generation—including at times when the region needs it most.

In the case of SunZia, the “how” includes a whole lot of patience. The project involves a remarkable 916 wind turbines in New Mexico and a 550-mile transmission line to carry the power to Arizona, where it goes on to serve customers in California. And it took over two decades, most of that in permitting, to get to this stage.

Even as it is coming online, SunZia is breaking other records: the amount of wind generation showing up in California’s electricity mix has leapt to new records multiple times in the last few weeks.

What’s next: Overall, the way forward for wind power is much fuzzier and more challenging than it should be, in part because of the Trump administration’s particular animus toward it, especially offshore wind. So one thing to watch for is the progress of the various lawsuits against the administration, which are trying to put an end to their various illegal moves to block wind farms.

But some next steps are clearer, including for the handful of offshore wind projects that are moving forward.

• The workers responsible for bringing Vineyard Wind to life installed the blades on the final turbine last month, so once the electrical systems and final tests are done, it’ll be all systems go.
• Revolution Wind, also in the area south of Massachusetts/east of Long Island, is itself almost fully built, and just started sending power to New England too. That contribution will ramp up as the project moves to completion.
• The Coastal Virginia Offshore Wind project, under construction, will be the largest in the country and one of the largest in the world when it comes fully online next year; last month it sent its first megawatt-hours to the mid-Atlantic electricity grid from some of the turbines already in place.
• Two other offshore wind farms also under construction, Sunrise Wind and Empire Wind, will soon be strengthening the New York electric grid.

Onshore, the new SunZia wind and transmission capacity will continue to show up in increasingly high levels of wind for California.

And wind records will continue to be broken elsewhere. Each of the other regional transmission organizations (grid operators) across the country—in the Mid-Atlantic and Midwest, New York, the midcontinent, Texas, and the Southwest—have scored records in the last few months.

US renewable electricity passes the 25% mark

What happened: Renewable energy generated more than a quarter of US electricity in 2025 for the first time.                                                               

How it happened: All that new renewable energy capacity adds up to more renewable electricity. Wind maintained its spot as the top renewable source, accounting for more than 10% of US generation last year. Solar’s fast growth brought it to almost 9%—four times its contribution in 2018. Hydro power was the next largest contributor, at 5.5%.

What’s next: Though there’s been considerable variability over the years, renewable energy’s contribution to overall US electricity supply has increased by an average of 1.2 percentage points per year. The federal Energy Information Administration (EIA) forecasts that renewables will be the source of the most growth in generation in 2026, with solar increasing 17% and wind 5%.

US renewable electricity out-generates gas on a monthly basis

What happened: In March 2026, for the first time ever, US renewable electricity beat gas generation over the course of a full month. Renewables accounted for 35% of generation, vs. gas’s 34.4%, according to Canary Media.

How it happened: Spring is always a strong time for renewable energy—the sun is shining, the winds are blowing, the waters are flowing. And the interlude between winter’s cold and summer’s heat is when demand for electricity is lowest. That all means less demand for gas. This recent crossover milestone, though, is principally a consequence of all the new renewables capacity of late.

Renewables’ strong showing last spring meant that these sources plus nuclear generated more than half of US electricity in March 2025 for the first time ever. That is, fossil fuels fell below the 50% mark. That happened again in April 2025 and March 2026.

What’s next: Watch for more months when renewables outperform gas, and fossil fuels get pushed below 50% (as a prelude to much greater push-downs in the years ahead); spring, and then fall, will be where that phenomenon will be more likely. More such achievements will require renewables continuing to grow and outpace increases in demand at a time when demand for electricity is growing more rapidly than it has for years—and all while contending with a major push for new gas-fired power plants and a Trump administration aggressively thwarting new clean energy deployment. But renewables have proven themselves time and again.

More 2025 wins, around the world

Renewable energy is making even more of a mark globally. Here’s a sample:

• According to analyst firm Ember, wind and solar generated more power than fossil fuels in the European Union for the first time over the course of 2025.
• In Colombia, solar power alone supplied more electricity than coal on an annual basis for the first time.
• Solar globally grew by a record amount in 2025, said the International Energy Agency (IEA), with China accounting for more than half that growth, the European Union achieving record numbers, and India posting 60% growth.
• Solar was “the single largest contributor to growth in global energy supply in 2025,” the IEA said—”the first time on record that a modern renewable source has led global primary energy supply growth.” Solar supplied more than 25% of that increase, compared to 17% for gas.
• Wind installations globally also hit a record (also led by China).
• Overall, the world installed a record amount of wind and solar—17% more than in 2024.

Lots more clean energy is on its way

Renewable energy’s momentum is a product of the many advantages it offers. Renewable energy is not just the cleanest but often the cheapest source of new electricity in the United States. Solar and wind have the advantage of potentially being faster to get installed than new gas plants, given the multi-year backlogs for gas turbines. And worldwide disruptions have brought into stark relief more of the risks of dependence on fossil fuels.

None of that means that new fossil fuel plants won’t get built, particularly given a big thumb on the scales from the White House. It does mean, though, that where clean energy is allowed to compete, the outcomes will likely to continue to testify to those advantages.

And projections bear that out. Solar, wind, and energy storage (batteries) combined had a record year in 2025, and made up more than 90% of new energy capacity in this country, according to industry association American Clean Power. And they will make up 93% of what gets built in the power sector in 2026, forecasts the EIA. All that will add up to more clean energy, and give us even more possibilities for phasing down fossil fuels and accelerating the transition to a clean and just energy future.

The transition to clean energy is a marathon, not a sprint. The people and communities behind clean energy in its many forms and uses will continue to push boundaries and break record after record. Count on it.

Operators botched a safety test and took the reactor into an unstable state, causing a rapid rise in power that triggered violent steam explosions that blew apart the reactor core and surrounding structures. Fires burned for days. A massive amount of radioactivity dispersed across the former Soviet Union and much of Europe. Hundreds of thousands of individuals were evacuated or relocated from contaminated areas, and a 30-kilometer radius “exclusion zone” was established that is still in place today. Dozens of emergency personnel died within weeks from acute radiation syndrome, and thousands of children developed thyroid cancer from radioactive iodine exposure. Ultimately, tens of thousands of cancer cases throughout Europe are projected to occur from the radioactive pollution caused by the disaster.

The United States and many other countries have sought to distance themselves from the potential for a Chernobyl-like accident by asserting that their nuclear regulators would never have licensed a reactor with the safety flaws of the RBMK (a Russian acronym for “high-power channel-type reactor,” the Chernobyl-4 design), and that light-water reactors (LWRs), by far the most common type of power reactor in operation, are far safer. While this argument has some validity, soon after the accident it became clear the safety benefits of LWRs compared to the Chernobyl-4 RBMK were not as great as advertised—a point later illustrated by the 2011 Fukushima Daiichi triple LWR meltdown in Japan. And today, many of the regulatory requirements and standards that underlie this confidence in the safety of the US nuclear fleet are being thrown by the wayside as the Trump administration recklessly pushes to “unleash” nuclear energy as quickly as possible.

It can’t happen here… or can it?

After the April 1986 Chernobyl accident, the US Nuclear Regulatory Commission (NRC), the independent nuclear safety and security agency created a little over a decade earlier to oversee commercial nuclear facilities, conducted a review of its potential implications for the safety of US nuclear power plants. In the process, the NRC convinced itself that such a catastrophe could simply not happen here. The agency highlighted numerous factors that distinguished the Soviet approach to nuclear power plant design and operation from that of the United States and other Western countries. Chief among these were requirements for nearly leak-tight, robust “containment” structures, strict limits on operation in unstable states that could experience rapid, uncontrollable power increases, and offsite emergency plans to protect the public in the event of a serious accident. Consequently, the NRC concluded that “no immediate changes are needed in the NRC’s regulations regarding the design or operation of U.S. commercial nuclear reactors.”

While the NRC did not believe its regulations were inadequate after Chernobyl, it certainly didn’t suggest they were too tough at the time. But in subsequent decades, the agency has been under constant pressure from the nuclear industry and many lawmakers to weaken its standards. This resulted in the lax oversight that allowed the Davis-Besse reactor in Ohio to come close to experiencing a serious loss-of-coolant accident and potential meltdown in 2002, and forced the NRC to temporarily slow down the pace of deregulation. But ultimately, the industry influence campaign culminated in the executive orders signed by President Donald Trump in May 2025, which have undermined the foundation of independent nuclear facility licensing and oversight that Congress put in place over 50 years ago when it split the Atomic Energy Commission into separate regulatory and promotional agencies—the present-day NRC and DOE.

And the worst is yet to come. The NRC is in the process of rewriting all its regulations and guidance in response to EO 14300, “Reform of the Nuclear Regulatory Commission,” with the explicit purpose of watering down safety and security standards to accelerate licensing of new facilities and reduce oversight of operating ones. And EO 14301, “Reforming Nuclear Reactor Testing at the Department of Energy,” directed the DOE to create a pilot program that would expedite the approval of three new nuclear reactors with the goal of achieving “criticality” (initiating a neutron chain reaction) by July 4, 2026—requiring an unprecedented and reckless rate of speed for construction and commissioning.

As a result of this vast lessons-unlearned exercise, companies may soon be building reactors across the United States that have more in common with Chernobyl than most people may realize. For example, in March, the NRC issued a permit in record time to TerraPower, a company co-founded by Bill Gates, to construct a 345-megawatt power reactor called the “Natrium” in the town of Kemmerer, Wyoming. The NRC approved the Natrium, a fast-neutron reactor, even though the design lacks a containment structure, is vulnerable to rapid, autocatalytic power increases, and uses a coolant, liquid sodium, that can catch fire—all adding up to what can be reasonably called a “Cowboy Chernobyl.”

Chernobyl lessons: learned and unlearned

In 1989, following its review of the causes of the Chernobyl accident, the NRC issued a report entitled “Implications of the Accident at Chernobyl for Safety Regulation of Commercial Nuclear Power Plants in the United States.” The report listed a number of specific areas in response to the accident that warranted attention by the NRC. These included reactivity accidents, accidents at low and zero power, multiple-unit protection, fires, containment, emergency planning, severe-accident phenomena, and graphite-moderated reactors. Today, the NRC, the DOE, and the nuclear industry are all busy unlearning the lessons of Chernobyl in each of these areas. Below, we focus on one of the most critical: the need for containment.

To contain (with a structure) or not to contain (with a structure)? That is the question

Ten years ago, on the 30^th anniversary of Chernobyl, the NRC once again invoked the critical role of reactor containments in differentiating the safety of the US fleet from the Chernobyl design:  “U.S. reactors have containment buildings equipped with walls that are several feet thick and have a steel liner on the interior to help prevent the release of radioactivity during a severe accident,” NRC spokesman Neil Sheehan said in a statement. “During the Three Mile Island Unit 2 accident, the containment structure served that function effectively.”

Yet only two years later, the NRC decided to abandon its longstanding design principle that “reactor containment and associated systems shall be provided to establish an essentially leak-tight barrier against the uncontrolled release of radioactivity to the environment …”. For new, non-light-water reactors—designs like the Natrium that use coolants other than ordinary water—the agency gave the green light for approving designs without conventional containment structures. Instead, the NRC would also accept so-called “functional” containments, wonk-speak for a regulatory rollback that would allow reactor applicants to take credit for other design features to provide a containment-like function and forgo a physical containment.

In addition to the Natrium, another proposed non-light-water power reactor design without a containment is the Xe-100, an 80-megawatt, “pebble-bed” high-temperature gas-cooled reactor (HTGR). Its developer, X-energy, is applying to the NRC to build four Xe-100s adjacent to a Dow chemical plant in Seadrift, Texas. Unlike LWRs, these types of reactors use “tri-structural isotropic” (TRISO) fuel,  which the DOE likes to say is “the most robust fuel on Earth.” X-energy and other HTGR developers claim that the fuel is so safe that a physical containment is not needed.

However, as detailed in this legal filing that UCS helped prepare for a hearing petition filed by the group San Antonio Bay Estuarine Waterkeeper against the Xe-100 construction permit application, TRISO fuel is not nearly as robust as its promoters claim, and there is significant uncertainty about how it will perform in certain types of accidents. Indeed, X-energy’s own accident analysis shows that the fuel could exceed the maximum safe temperature by several hundred degrees Celsius. The company has simply not made the case at this preliminary stage that the reactor can be safely operated without a containment.

X-Energy has only recently begun a multi-year testing program to attempt to address uncertainties in fuel performance. Nevertheless, the NRC appears poised to allow construction of the containment-less design to proceed before the data from such testing is obtained. But even if the fuel performs far worse than the application assumes, it is highly unlikely that the NRC would require X-energy to undertake an extremely costly retrofit to add containments to the four reactors before allowing them to operate.

Given the emphasis that the NRC formerly placed on the safety benefits of physical containments, why would it take this step? The answer is simple: cost. Physical containment structures, which typically require large quantities of high-quality reinforced concrete, are expensive. A 2020 MIT study found that the containment was one of the largest contributors to the cost of light-water reactors. Thus a quick way to cut nuclear power project costs would be to leave out the containment.  And for some new reactor types and deployment models—think small modular reactors being hauled by truck to your local data center—having a physical containment would be completely impractical as well as unaffordable. 

The problem here is that eliminating physical containments is a truly pound-foolish approach to reducing the high cost of nuclear power, as the Chernobyl experience has shown. What the NRC is allowing them to be replaced with, functional containment, is not an adequate substitute. One of the primary roles of containment is to provide “defense-in-depth”—an extra level of assurance in place to compensate for gaps in understanding of how the reactors themselves will work during accidents. And for new reactor designs with limited or no operating experience, there is an awful lot that the developers and the NRC simply do not understand and cannot accurately predict. Thus, the role of containment in helping to offset the unknown risks posed by new, experimental designs is more important than ever.

But the NRC is approving functional containments for new reactors, such as the Natrium and the Xe-100, based on paper safety studies that have had little to no actual real-world validation. The agency is allowing applicants to exclude accident scenarios from their safety analyses that could demonstrate the need for a physical containment, based on speculation that they are so improbable they do not need to be considered. This is little different than the approach Soviet reviewers took when they approved the Chernobyl design with only a partial containment.

According to the NRC’s 1986 Chernobyl review, “credible accidents with potentially serious consequences” were not discussed in the Soviet safety analysis, “presumably because they [were] considered to be of sufficiently low probability to justify disregarding them in the design basis.” These included “rapid reactivity excursions” and other accident sequences that occurred during the Chernobyl accident. The report also stated that “an important result of the decision to consider only pipe breaks below the reactor as credible is that there is no containment surrounding the outlet piping above the reactor.”

This reasoning will sound familiar to anyone acquainted with the “risk-informed” regulatory approach that the NRC  recently approved for new reactor licensing, which includes processes for addressing questions such as the adequacy of functional containment. In the new 10 CFR Part 53 rule issued in March, applicants are allowed to use either “probabilistic risk assessments” or “systematic risk evaluations” to develop the spectrum of accident sequences that are considered in the licensing basis. The lack of any guidance for carrying out such a “systematic risk evaluation,” including a standard for determining the worst-case accident that needs to be considered in designing a reactor, provides ample opportunity for the kind of subjective cherry-picking that the Soviets used in developing the Chernobyl design and justifying the lack of a full-blown containment.

The NRC’s approval of the Natrium construction permit and its likely approval of the Xe-100 are setting dangerous precedents for all future reactor proposals, most of which are containment-free designs. But fortunately, there has been very little actual new nuclear plant construction yet, and the NRC has ample opportunity to change course before it allows irreversible mistakes to be made. On the occasion of the 40^th Chernobyl anniversary, the NRC should take a hard look in the mirror and reconnect with its 1989 finding that “the most important lesson [of Chernobyl] is that it reminds us of the continuing importance of safe design in both concept and implementation … and of backup features of defense in depth against potential accidents.”

This week also marks the 20^th anniversary of my own visit to see the devastation within the Chernobyl exclusion zone, which is seared into my memory. If anyone still needs convincing that it’s a terrible idea to allow reactors without real containment structures to be built across the United States, I highly recommend the Chernobyl tour (once the war is over of course).

One of the barriers slowing down clean energy development has been long wait times to connect projects to the grid. The interconnection process is how generating projects receive approval from California’s grid operator (CAISO) to connect to the grid. CAISO runs a series of studies to determine whether a project can safely connect to the grid, or if additional grid upgrades are needed to accommodate the new project.

The interconnection process is a critical step to getting new clean energy resources online. However, it had gotten bogged down by an overwhelming number of applications and become a notable barrier to connecting new clean energy. This blog post will cover CAISO’s reforms to the interconnection process and their effectiveness at addressing the long wait times.

How did the interconnection process work prior to reforms?

CAISO used to run interconnection studies for all projects submitted to them. Interconnection studies are tricky because the ability of a project to connect to the grid also depends on the other generating projects already connected, and planning to connect, to the grid. For example, if a planned project drops out, the studies that incorporated that project become less precise. This dependency makes interconnection studies more complicated as more projects are being studied. Although the time to get through the interconnection process was slowing down, the number of projects being submitted was manageable for the process to move along.

In 2023, CAISO decided that was no longer the case. During its annual application window, which are named clusters, with 2023 being Cluster 15, CAISO received 347 gigawatts (GW) of interconnection requests from 541 projects, compared to 373 projects in Cluster 14 and 155 in Cluster 13. This large spike in Cluster 15 was on top of 185 GW already sitting in the interconnection queue. For perspective, CAISO also noted that only 165 GW of new resources would be needed to meet the state’s 2045 clean energy portfolio.

What reforms were implemented?

To address the interconnection issue, CAISO initiated reforms, implementing an Interconnection Process Enhancement (IPE) 3.0 in 2023 with three steps to improve the process. First, CAISO extended the deadlines for Cluster 14 studies, and paused Cluster 15 studies to give the agency time to reform the interconnection process for Cluster 15 requests and beyond. The next step was focusing on these new reforms.

At a high level, the reforms were intended to limit the number of projects being studied and prioritize projects to align with grid needs, market interest, and project viability. Specifically:

• CAISO now accepts only enough projects to meet 150 percent of the available capacity based on the point of interconnection’s transmission constraints. This limits the number of projects being studied to a more manageable number.
• All projects requesting to connect to the grid are now scored and ranked to decide which projects pass the 150 percent capacity threshold to reach the study stage. This prioritizes the projects being studied.

The scoring system uses metrics across commercial interest, project viability, and system need to rank projects with indicators in each of these categories below. Commercial interest points are given by load-serving entities (LSE), and other large electricity buyers and project viability are provided by project engineers.

Projects with the highest scores at each point of interconnection move to the next stage up to the 150 percent limit. As a note, because the 150 percent capacity is based on transmission constraints at each point of interconnection location, projects with the highest overall scores may not necessarily move forward, since some points of interconnection are more competitive than others.

The final step of IPE 3.0 modifies CAISO’s transmission plan deliverability (TPD). The TPD process allocates deliverability capacity for projects connecting to the grid. Projects must secure deliverability capacity to be eligible for resource adequacy, which is often important for project developers to secure project financing and buyers. Updating the TPD process was needed to align it with the other reforms and ensure that the most viable projects received deliverability capacity.

Has the new initiative been effective?

With the final step of IPE 3.0 approved only earlier last year, it’s difficult to fully assess the impact of the reforms. Broadly, the goals outlined at the outset—reducing project intake and queue management—have been successfully addressed. Cluster 15 has been reduced to a significantly more manageable number of projects that will go through the interconnection studies, and these projects are seemingly more viable by being further along in the development process. When the application window was initially opened in 2023, 541 projects totaling 347 GW were submitted. Following the implementation of the new process, 145 projects totaling 68 GW were passed to the stage for interconnection studies.

Cluster 15 is a good initial case study for the reforms. While generally a smooth process, there are a few notable issues that stand out:

1. Cluster 15 reveals the high variation of project scores that pass into the validation stage. Low scores were able to move to the next stage in some areas, whereas high scores did not pass in other areas depending on how competitive the area was and how much capacity was available.
   • Recommendation: More geographically granular data on transmission constraints would be useful for informing where projects could be sited. This would ensure that projects with higher scores, and are thus more viable, are being moved forward across the grid.
2. A notable number of projects were withdrawn after passing the 150 percent threshold stage. The number of projects decreased from 177 to 145 with an associated decrease of almost 30 percent of capacity. Even with the increased likelihood that these projects are completed, additional withdrawals could create concerns that CAISO is no longer moving enough projects forward.
   • Recommendation: The CAISO should consider another assessment window after projects are withdrawn to allow the next highest scoring projects to move through.
3. A potential deficiency in the process is an equity component to support environmental justice objectives. The scoring system does not consider and could even systemically disadvantage projects that serve these communities. For example, a recent petition at the Federal Energy Regulatory Commission noted that Tribes face more financial challenges to obtaining the high commercial readiness deposits in the interconnection process. The scoring and ranking process under IPE 3.0 could systematically exclude projects that may not be as financially viable, but serve critical grid reliability needs of Tribal or disadvantaged communities.
   • Recommendation: Future IPEs should implement pathways to ensure equitable access in the interconnection process for all communities to access clean energy.

The IPE is an iterative process, and many of the initial reforms in IPE 3.0 triggered additional issues that need to be discussed. The next iteration, IPE 5.0, launched this past year to address new and lingering issues such as managing stagnant projects in the queue. This continued engagement is important to ensure the processes for connecting clean energy generation to the grid are evolving to match the clean energy needs of the state.

Across the country, grid operators are struggling to complete interconnection studies in a timely manner, and connect clean energy projects to the grid. With the current federal administration hostile to clean energy, states like California need to remain steadfast in the transition to a clean grid. CAISO’s interconnection reforms are an important step in accelerating that transition.

While it may not feel like Louisiana is teeming with data centers just yet, the boom in energy-hungry artificial intelligence is poised to change the landscape. We’re talking about multiple cities’ worth of electricity demand being added to the grid over the coming decade.

New modeling by the Union of Concerned Scientists has found that data center growth could leave Louisianans paying for billions of dollars in additional electricity system costs over the next 15 years. And under current policies, the AI facilities in the state are set to be powered largely by fossil fuels, bringing potentially billions of dollars in public health costs and tens of billions in global climate damages.

Preparation for this type of massive, yet highly uncertain, load growth requires careful attention by regulators and policymakers tasked with protecting the public. In other parts of the country, data centers have brought risks of costly and dangerous power outages while also raising utility bills at a time when energy costs are already rising for several other reasons. And depending on how data centers are powered, they can bring significant harms to public health and the global climate. Unfortunately, Louisiana’s current policies and regulatory approaches are not well set up to address the wide array of risks posed by the data center boom.

Fortunately, there are steps that policymakers and regulators, particularly the staff and elected officials at the Louisiana Public Service Commission (LPSC), can take to protect their constituents from these risks and ensure that Big Tech’s burdens don’t fall on Louisiana residents and businesses. Let’s get into the details.

Data centers set to make Louisiana’s grid way more expensive

Depending on the extent of data center load growth, our findings show that over the next 15 years, Louisiana’s wholesale electricity system costs could be a cumulative $14 billion to $26 billion higher than they would be without data center growth. We call these the “Mid” and “High” data center growth scenarios, respectively. This analysis draws from state-level results from our Data Center Power Play report, a national-level study using the Regional Energy Deployment System (ReEDS) modeling framework that was released earlier this year.

Louisiana ratepayers are at risk of paying substantial electricity system costs caused by data centers. “Bulk” electricity system costs are only at the wholesale level. Calculation was done by comparing the Mid and High Data Center Growth scenarios with a No Data Center Growth counterfactual scenario. Source: UCS

These costs are only at the wholesale level—essentially, the costs to build and operate large-scale power plants and transmission lines. It doesn’t reflect ”ratemaking” at the LPSC, the process whereby those wholesale costs are allocated to residents and other businesses.

However, the dollar amount reflected on energy bills includes other costs as well, such as the utility company’s profit margin.  These additional costs are covered by retail ratepayers, like residents and businesses. And Louisiana does not have comprehensive protections to insulate ratepayers from data center-triggered costs. In fact, the LPSC’s recent fast-track approval pathway, established through the recent “Lightning Amendment,” clears the way for potentially more than half of such costs to be passed to other ratepayers.

Data centers’ projected impact on the average Louisiana utility bill is uncertain, because that depends so heavily on how the LPSC allocates the wholesale electricity system costs between different types of customers (e.g., residential, commercial, industrial). But with electricity system costs potentially $26 billion higher due to data center load, and without comprehensive protections in place for other ratepayers, Louisianans are at risk of substantially subsidizing—to the tune of billions of dollars—Big Tech’s AI ventures.

Status quo would keep Louisiana over-reliant on a single fossil fuel: gas

About 75% of Louisiana’s electricity generation is currently from fossil gas power plants, making it one of the most gas-reliant states in the nation. Our analysis shows that under current policies, the state will meet growing demand with even more gas. This includes demand from data center companies, which thus far have not made any attempt to plan for flexible operations (essentially reducing demand during times of grid stress) in an effort to reduce overall costs and the need for new fossil fuel plants.

Therefore, without policy changes, the Louisiana power grid’s overdependence on a single fossil fuel, gas, would sustain into at least the 2040s, making up roughly two-thirds of the electricity mix in our 2041 modeling results.

Under current policies, Louisiana is projected to stay overreliant on gas-fired electricity. Source: UCS

The unpredictable spikes in utility bills that Louisianans are all-too-familiar with would therefore continue, since utilities pass fuel cost increases directly to their customers. The latest spike was caused by Winter Storm Fern in January 2026, which sent gas prices soaring above $30 per million British Thermal Units (MMBtu)—the highest in at least 29 years. For perspective, the price was around $3 per MMBtu just a week earlier. Though the effect on utility bills is not yet clear, ratepayers will feel the impacts of those price increases in the coming months even if they use the same amount of power.

While short-term commodity price changes aren’t captured by long-term modeling frameworks like ReEDS, those spikes can still have significant real-world impacts on energy burdens.  Some Louisianans were paying bills in 2025 that were 29% higher than the year before due to increases in gas prices. And during a 2022 price spike, some customers were paying double the fuel charges—these days roughly 20-30% of a total bill—than they were paying the year before.

Diversifying away from price-volatile fossil fuels and toward zero-marginal-cost resources like wind and solar can help protect ratepayers from these types of bill increases. Otherwise, Louisiana households will continue to be forced to fund the unpredictable costs of utilities’ overreliance on those fuels, whose price is sensitive to an increasing number of extreme weather events, and global conflicts such as the wars in Ukraine and Iran. While the US has thus far been insulated from the latter in terms of gas prices, that is not all guaranteed to be the case as the war continues.

Data centers set to bring higher public health and climate damages           

Beyond utility bill increases, data centers are also set to trigger higher public health costs and climate damages from Louisiana’s gas plants. Our findings show that the public health damages could range from $1.5 billion to $3 billion from 2026-2041 due to increases in nitrogen oxides (NO_x) and sulfur dioxide (SO₂) emissions, two pollutants that can cause respiratory and cardiac issues.  While these public health harms can cross state lines as pollutants flow downwind, the impacts are predominantly local.

Data centers drive billions of dollars in public health and climate damages as Louisiana relies on gas plants to meet growing electricity demand. Calculation was done by comparing the Mid and High Data Center Growth scenarios with a No Data Center Growth counterfactual scenario. Source: UCS

Over the same period, data center-driven increases in heat-trapping emissions from Louisiana fossil fuel plants could trigger $35 billion to $87 billion in global climate damages. While these damages are felt globally,  Louisiana already experiences a number of impacts that scientists expect to worsen as climate change continues, including hurricanes, heat waves, and sea level rise. It is therefore imperative that the state make concerted efforts to reduce both toxic air pollution like NO_x and SO₂, as well as heat-trapping emissions like carbon dioxide and methane.

But how much data center demand will actually come online?

One big question remains: how much data center growth is actually coming to Louisiana? The short answer: no one knows.

Regulated utilities have financial incentives to overestimate demand and overbuild, because they earn ratepayer-funded profits on capital infrastructure spending. We therefore have to take data center demand estimates from utilities with a skeptical eye.  

To account for the uncertainty, UCS ran multiple data center demand scenarios at the national level. Our “Mid Growth” scenario is in the range of other national-level studies. (See this blog for more on our latest national data center analysis.) However, in Louisiana specifically, looking at recent announcements in a vacuum makes the “High Growth” scenario seem far more likely, and maybe even conservative.

Our High Growth scenario projects about 5 GW of data center load added to Louisiana’s grid by 2041. Let’s compare this to Meta Platform’s plans for a new data center near Rayville, LA. The size of the data center expansion is thus far confidential, but Mark Zuckerburg said last year that the facility could grow to 5 GW, which would consume roughly six times the electricity as the entire city of New Orleans on an annual basis. Last year, the LPSC approved Entergy Louisiana’s application to build 2.3 gigawatts (GW) of gas capacity for this data center. And Entergy recently filed another LPSC application to build seven new gas plants totaling 5.2 GW on top of the already approved 2.3 GW, all for the expansion of Meta’s data center.

There’s much more to be said about this new application and who will end up covering the costs. But for now, I want to underscore the significant remaining uncertainty with the data center landscape in Louisiana and beyond, even as the press releases and LPSC applications make it all seem like a foregone conclusion.

To understand the uncertainty, we don’t have to look any farther than Meta itself. Right after Entergy got approval to build the first 2.3 GW of gas capacity for the data center, the tech giant fundamentally changed the financial structure of the planned AI facility. Meta offloaded 80% of the data center project ownership onto Blue Owl Capital—a much riskier company—and gave itself the option to exit the data center lease after just four years. The electricity infrastructure being built, meanwhile, will last for decades. Meta has financially shielded itself greatly, in no small part by getting a ratepayer guarantee of this long-lasting infrastructure.

Worries about an AI bubble bursting have only grown since UCS conducted this modeling in late 2025. These worries are due to a number of factors, including circular financing, lack of AI profitability in comparison to massive capital expenditures, private credit scares (of which Blue Owl is at the center), and now, Donald Trump’s war in Iran.

If the AI bubble isn’t bursting in the way some warn, and our High Growth scenario proves to be on the conservative side, then the need for safeguards is even more urgent, because the impacts will be that much greater. Policymakers at the LPSC should act now to protect communities from the wide array of risks stemming from the growth in data centers.

Looking ahead: better protections from data center threats are needed

As discussed above, we estimate that the growth in data centers could cause up to $26 billion in additional Louisiana electricity system costs between 2026 and 2041. The state would remain alarmingly overdependent on gas for its power sector needs, leaving ratepayers highly vulnerable to unpredictable price shocks. The additional pollution from these fossil fuel power plants specifically to serve data centers would trigger up to $3 billion in public health damages and up to $87 billion in global climate damages.

Fortunately, the commissioners at the LPSC have the ability to stave off a situation which is untenable for many of their constituents, particularly since an estimated 50% of households in the state are already financially struggling.

We recommend several reforms in our issue brief that would begin to put the state on a path toward a cleaner, more affordable, and more reliable electricity system. Included in these recommendations is an improved process for long-term utility resource planning, as well as comprehensive, mandatory ratepayer protections from data center-triggered costs.

We also recommend that the state take advantage of its significant clean resource potential, including by embracing long-range transmission planning by regional transmission grid operators MISO and SPP. Further, reforms are needed to enable a wider set of stakeholder voices to inform decisionmaking at the LPSC. For far too long, utilities have had disproportionate influence at the agency, and that is being perpetuated in part by low transparency and arbitrary barriers to participation.

The time is now for Louisiana utility regulators to protect their constituents from data center threats, and you can urge them to do so at this link. They should not continue to cater to Big Tech and utility company interests at communities’ expense.

Repealing the clean energy standards would be a step backward, failing to solve reliability or cost concerns while ignoring the real emerging challenge: load growth from AI data centers. In fact, UCS analysis of this load growth demonstrates that to ensure a clean, affordable, and reliable energy future, Michigan must strengthen the clean energy standards—not abandon them. It also reveals loopholes that would increase fossil fuel use, even without the repeal.

In this blog, I’m going to break down the structure of the current clean energy standards in Michigan, demonstrate how load growth from data centers puts them to the test, and explain how they must be strengthened to keep moving toward a clean, healthy, and affordable energy future.

The basics of Michigan’s clean energy standards

There are three key points to keep in mind about the clean energy standards:

The categories of energy sources overlap. Wind, solar, and hydro-power are included in the definition of “renewable energy,” while “clean energy” includes renewables in addition to other low carbon energy sources, such as nuclear and fossil gas with carbon capture and storage (CCS) (see Figure 1). While labeling some of these sources as “clean” is a bit of a misnomer,  I’ll refer to them as clean here to be consistent with the Michigan standards.

Michigan’s energy standards phase in over time. The renewable energy requirement started at 15% in 2024, increases to 50% in 2030, before reaching 60% in 2035. The clean energy requirement starts at 80% in 2035 then jumps to 100% in 2040.

The standards only apply to retail electric sales to “end users” in the state, while also allowing for renewable and clean energy to be purchased outside of Michigan.

Despite the clean energy standards, Michigan’s emissions increase

The Union of Concerned Scientists recently published an analysis of potential load growth from data centers in the United States, demonstrating how that growth affects the grid under different policy pathways. We included a deep dive in Michigan, which produced a surprising result: despite the clean energy standards, heat-trapping emissions from power plants increase steadily over time (see Figure 2). 

This effect appears across all load scenarios, but because data centers drive massive new demand, they act as a “stress test” for Michigan’s clean energy standards. Our analysis shows that while the standards reduce emissions early on, emissions begin rising steadily after about 2035.

A look at the underlying generation mix helps explain how this is possible. Figure 3 shows the shares of energy generation across renewables, other clean energy, and fossil fuel (coal and fossil gas without CCS) under the “current policies” scenario in our analysis.

This highlights the problem: shouldn’t renewables plus other clean sources hit 100% by 2040? Let’s explore why that isn’t the case in our modeling results.

Clean energy standards don’t apply to exports

The first big caveat to Michigan’s clean energy standards is that they only apply to energy sales within the state: any electricity that is exported elsewhere is exempt. According to our analysis, exports increase dramatically over time (see Figure 4).

These exports are not subject to the clean and renewable energy standards. This “export loophole” allows utilities to cover their retail sales with clean and renewable energy, while fossil plants can continue running and exporting the energy they produce without limits.

As demand grows due to data centers, our modeling shows that this loophole becomes more consequential, with exports increasing steadily under current policies. But under a different clean energy pathway (which I’ll detail below), exports are subject to clean energy requirements, meaning they don’t cause emissions to increase.

While that growth in exports is significant, it’s still not enough to explain why carbon emissions are increasing over time.

Utilities can buy clean and renewable energy if they don’t make enough themselves

The next caveat is that Michigan utilities can buy credit for renewable and clean energy that they don’t generate themselves. This is where things start to get complicated. Utilities have three basic pathways to meet the state’s clean and renewable standards:

• Generate clean/renewable energy themselves
• Purchase clean/renewable energy directly
• Purchase clean/renewable energy attributes

For renewable energy, the second pathway is often referred to as “bundled renewable energy credits” (or “bundled RECs”), where a utility both buys the energy from a renewable generator and takes credit for its renewable attributes. Typically, this means the utility is paying for the energy from a specific renewable energy project.

The third pathway is “unbundled RECs,” where a utility only pays for the renewable attributes, without the associated energy.

To oversimplify, a utility buying bundled RECs is replacing some of the energy it would sell to its customers with renewable energy generated somewhere else. But a utility buying unbundled RECs is continuing to sell dirty energy to its customers, and paying what amounts to a fine. Unbundled RECs are the indulgences of the energy world.

Fortunately, Michigan has limits on how many unbundled RECs can be used to meet the clean energy standards; they have to come from within the same regional markets that Michigan is a part of (MISO or PJM, depending on the utility), cannot exceed 5% of the utility’s total, and can’t be used for compliance after 2035.

While these limits are a good thing, and bundled RECs are certainly preferable, there’s still a big caveat: Since utilities can just purchase bundled RECs to cover their obligations for retail sales within their territory, there’s nothing stopping power plants from continuing to generate dirty electricity to sell somewhere else.

How does Michigan meet the standards while fossil generation increases?

Now that we understand some of the loopholes in the clean energy standards, let’s take a look at how Michigan is actually meeting these requirements in our modeling, despite the rising emissions. Figure 5 shows the mix of in-state and imported energy credits that are used to meet each of the requirements in 2035 and 2050.

In 2050, Michigan relies on significant imports of clean energy from other states to meet the 100% clean energy requirement which takes effect in 2040, even meeting the requirement ahead of schedule in 2035. The shares for clean energy exceed 100% for a few reasons, but mostly due to the fact that Michigan generates more energy than it needs, leading to exports to other states.

Due to some nuance in how the model calculates and reports renewables requirements, the figure doesn’t quite show Michigan hitting the 60% requirement that takes effect in 2035, but we can see that imports here are also required to comply.

In contrast to the mix of energy credits used to comply with the standard, Figure 6 shows what the shares of actual generation (including exports) look like.

The results show a system where legal compliance and physical reality diverge. Michigan is not actually decarbonizing its energy supply, and it’s getting worse over time—even as it complies with the clean energy standards. Most glaring from our results is that in 2050, Michigan would be “meeting” its clean energy standards while nearly 60% of energy generated in the state comes from fossil fuels.

A better path: regulating emissions directly

To explore more robust clean energy policies, we proposed an alternative scenario based around a CO₂ Reduction Policy in our data center load growth analysis.

The key feature of this policy is that it regulates actual power plant emissions, including both imported and exported energy. This approach closes the loopholes in the existing laws: utilities can continue to sell to (or buy from) other states, but all of that energy is subject to emissions limits, not just the portion that’s sold to end users in the state.

Table 1 shows how we modeled phasing the policy in over time, compared to the existing renewable and clean energy requirements. Though the fossil fuel industry is likely to object sharply, this phased approach gives utilities a reasonable planning timeline for compliance.

You can see the impact in Figures 7 and 8. First, the generation plot shows that renewable and other clean energy sources increase their share over time, approaching 97% of the total by 2050.

Finally, Figure 8 shows that across all load scenarios, CO₂ emissions decrease over time, eventually going below zero due to technologies like biopower with CCS (which explains the last 3% from Figure 7).

Looking ahead

As data centers add enormous new load to the grid, significant new generation will be needed. The existing clean energy standards place guardrails on what types of generation are allowed, guiding the state towards a clean and renewable future. But as we’ve seen, the loopholes in these standards allow utilities and power producers to rely on accounting mechanisms to meet the letter of the law while still expanding fossil fuel generation.

This has measurable negative impacts: under the existing laws, our analysis shows that the expected growth in data centers would lead to $118 billion in climate damages and $1.6 billion in health damages by 2050 due to air pollution and emissions from fossil fuel power plants in Michigan.

Michigan’s clean energy legislation in 2023 was a strong start, but as load growth from data centers reshapes the system, stronger policies are needed. Rather than moving backward with misguided attempts to repeal the clean energy standards, they must be strengthened with actual limits, regulating emissions of heat-trapping gases directly. Policies that focus on actual emissions ensure that every megawatt-hour of electricity moves the system closer to the clean, healthy energy future Michiganders deserve.

But it’s not just what’s happening right now that’s deeply concerning—it’s also what happens next.

Because while today’s affordability challenges require urgent solutions, the ratcheting onslaught of power-hungry data centers is threatening to turn these challenges into a runaway crisis.

While decisionmakers at the state and local level are racing to navigate these issues, the Trump administration is actively exacerbating the problem, sidelining solutions while super-charging stressors across all the major points of friction.

In particular, the Trump administration is increasing demand for electricity at the exact same time as it’s constraining the availability of new electricity supply, resulting in a surging—and costly—supply-demand mismatch. But that’s not all. The Trump administration is also inflaming other major electricity cost drivers, including investments required for updating faltering grid infrastructure, preparing the electricity system for and helping it recover from increasing extreme weather events, covering rising natural gas prices, and managing market instability.

As a result, electricity costs aren’t just high: because of the antagonizing policies of the Trump administration, for many, they’re set to soar.

The Trump administration is super-charging the electricity supply-demand mismatch

An unprecedented rapid surge in electricity demand, overwhelmingly the result of an AI-driven build out of data centers, is colliding with a steadily stacking set of constraints on new electricity supply with no end in sight. Without intervention, the result is a system on the brink: spiking electricity prices, limits on expanded use of electricity, and increasing reliance on costly, decrepit, and highly polluting power plants well overdue for retirement.

Unwinding the supply-demand mismatch requires action on both the demand and supply sides of the ledger.

Accelerating the surge in electricity demand

For the demand side of the challenge, there’s what to do about data centers, and then there’s what to do about everything else. While “everything else” can be well-addressed by prioritizing investments in efficient and flexible use of electricity, data centers are posing novel challenges to the system: their electricity needs are absolutely massive, individually equivalent in use to entire cities, and they’re being erected seemingly overnight. Meanwhile, the rules and regulations that govern electricity supply, use, and costs were designed for gradual, incremental change.

In the gap, everyday electricity users—aka people and their monthly bills—have been left exposed and are subsidizing this literal power grab three times over: first in having paid for the electricity infrastructure that tech companies are now cashing in on; second in suddenly having to pay billions and billions more for new infrastructure being erected only because of new data centers alongside higher overall system costs; and third in bearing the overwhelming costs and risks of stranded assets if the AI bubble implodes, while tech companies are positioning themselves to simply walk away.

To shift the costs and risks back onto the shoulders of the tech companies that are causing them, UCS’s latest report recommends data centers should be required to bring their own new clean electricity, directly or through contracts, to ensure supply is keeping up with demand; be flexible to avoid contributing to intermittent periods of grid stress and high costs; and pay for higher induced system costs as well as purpose-built infrastructure, to insulate the public from near-term bill spikes and, just as critically, from the risk of stranded assets down the line.

But here’s what the Trump administration is doing instead.

To start, the administration has slashed access to the very easiest demand-side solution in the book: energy efficiency. In the budget bill President Trump signed into law last July, incentives for improving the efficiency of new and existing buildings, as well as new appliances, were slashed to zero within the year. The Trump administration has also repeatedly attacked and undermined the capacity of the Energy Star program and run roughshod over energy efficiency standards at the Department of Energy (DOE), which in one fit of pique announced the rescission of 47 standards in just one day. Beyond being a staggeringly bad deal for consumers (see this tracker from the Institute for Policy Integrity for the exact cost to consumers from all these rollbacks, on the order of hundreds of millions of dollars of lost benefits per year, per rule), these actions truly beggar belief in this electricity-constrained environment.

And then there’s managing demand from data centers.

While state-level decisionmakers are directly contending with managing data center energy system impacts, with responses ranging from fully steamrolling everyday people to contemplating temporary moratoria on new data center additions, the Trump administration has now attempted to insert itself straight into the middle of the narrative—and is doing so by attempting to play it both ways.

On the one hand, recognizing growing backlash to data centers at the local level and surging energy affordability challenges more broadly, the Trump administration is newly touting proposals that seek to shift data center costs back onto tech companies. For example, the Trump administration brought some of the largest tech companies to the White House in March to sign a voluntary “ratepayer protection pledge”—which theoretically could help, but for the fact that there are absolutely zero actual policies or regulations attached, nor does the administration have any authority over the policies outlined. That is not an accident. The Trump administration is plainly angling for an optics win while doing absolutely nothing to protect everyday people’s costs.

In a similar story, DOE Secretary Wright directed the Federal Energy Regulatory Commission (FERC) to begin a rulemaking that would standardize the process for connecting large loads (read: data centers) to the transmission system, as well as consider cost allocation related to those interconnections—but the directive was overwhelmingly focused on how to clear a path for expedited data center development, not defend against price spikes.

Meanwhile, the White House’s general approach to AI development, such as its July 2025 AI Action Plan, December 2025 executive order, and repeated efforts to shortcut permitting requirements and pollution regulations for new data center developments, point to an administration actually looking to fast-track AI at just about any cost—including uninhibited buildout of data centers and all their associated electricity demand.

Restricting access to new electricity supply

To address the second half of bridging the supply-demand gap, resolving the supply side of the crunch,there’s one main fix: build, build, build. Build as much electricity generation as possible, as fast as possible, as cost-effectively as possible, and make it clean. Build power supplies and enabling infrastructure in ways that work for the needs of today and for the clear and pressing needs of the future—a future where, unquestionably, for economies to be competitive, economies must be clean.

The resources that check these boxes best are renewables, most prominently wind and solar, alongside energy storage. In the overwhelming number of cases, these resources are the cheapest, these resources are the quickest to deploy, these resources are reliable, and, critically, these resources are clean, meaning unlike new fossil fuel-fired generation, they will remain relevant and useful long into the future.

The second part of the supply solution can be counter-intuitive: let faltering power plants retire. If a power plant is too costly to run, too unreliable to be counted on, or too polluting to be safely operated, the power plant should retire. Propping up decrepit coal-fired power plants just wastes money, upends years-long multi-faceted planning, and gives a false sense of security while boxing out resources readily able to replace them.

Build what works and retire what doesn’t. Simple, right?

Except what the Trump administration is doing is exactly the reverse: it’s blocking entities from building what works, and it’s blocking entities from retiring what doesn’t.

Since Day 1, the Trump administration has been working to sideline wind and solar generation while pushing fossil fuels. Wind and solar are here, today, ready to roll—but the Trump administration is doing absolutely everything in its power to keep them offline. The country’s grid is on the precipice, and the administration is going full-out to sideline the overwhelming majority of the solution set.

Consider this. Since his term began, President Trump and/or his administration has:

• signed a law that:
  • singled out wind and solar for rapid loss of incentives,
  • abruptly ended incentives for rooftop solar,
  • pulled funding for community-level solar projects and financing for catalyzing clean energy investments, and
  • shifted DOE loans away from wind and solar projects;
• repeatedly attempted to cancel permitted, and even nearly completed, renewable energy projects without any defensible reason;
• issued blatantly discriminatory policies that attempt to straight-up block, directly or in effect,the permitting of wind and solar resources on federal—and even non-federal—lands;
• launched specious investigations into alleged “health harms” of offshore wind;
• inflicted costly, arbitrary, and constantly shifting tariffs on key electricity generation and infrastructure inputs and components;
• pulled tens of billions of dollars in funding from awarded—and contracted—projects and initiatives intended to deliver clean energy resources;
• forced out federal experts and slashed offices working on advancing clean energy deployment, even going so far as to strike “renewable energy” from the name of the world-renowned National Renewable Energy Laboratory;
• and on, and on, and on.

The costs of these attacks are significant, in direct energy bill effects, in jobs lost, in investments cancelled, in innovation ceded, in sectors crushed—and because of the way they’re disrupting the pipeline of new projects and progress, the worst of these impacts won’t be fully felt for years to come.

Throughout all of these attempts to block new renewables, the Trump administration has slashed basic polluter accountability for fossil fuel-fired power plants and actively, repeatedly intervened in the market to force coal plants slated for retirement to stay online—even though utilities, regulators, and electric grid operators all want them closed, even though keeping them open is costing consumers tens to hundreds of millions of dollars apiece, even though they were each closing for a reason.

The Trump administration’s energy policy agenda of blocking what works and forcing reliance on that which doesn’t is a catastrophe, through and through.

Additional factors are driving electricity bills higher—and the Trump administration is making each worse.

The Trump administration’s gratuitous exacerbation of the supply-demand mismatch presents a significant and surging threat to electricity costs. But it’s not the only action by the Trump administration contributing to higher energy bills. In particular, each of four major additional drivers behind energy affordability challenges are also getting sent in the wrong direction by the Trump administration:

• Exposure to rising and volatile fossil gas prices. Gas-fired electricity generation represented approximately 40 percent of the US electricity mix last year. In some parts of the country, the share is even higher. That makes electricity prices heavily impacted by the rising price of natural gas. The Trump administration is exacerbating this challenge on three fronts: first by attempting to push utilities to be even more dependent on gas; second by actively encouraging unmitigated growth in liquified natural gas exports, which is expected to drive the domestic price of gas far higher; and third by repeatedly sowing geopolitical instability in the very markets to which he is further linking the nation’s fossil gas supply.
• Market instability. Since the very start of his term, President Trump and his administration have unrelentingly attacked the very foundations of what establishes the US as a stable investment and innovation environment. Foremost among these is President Trump’s reckless and destabilizing war and escalating embrace of violence. In addition, consider this: the imposition of shocking, arbitrary, and constantly shifting tariffs and disruption of global markets; fueling inflation; attempting to cancel nearly completedprojects; yanking billions upon billions of contracted and obligatedfunds; stifling and attempting to wholly dismantle the world-leading US scientific enterprise; weaponizing and politicizing the application of laws and policies; raiding a new, multi-billion dollar, collaboratively established factory; quitting long-standing and highly productive international coordinating bodies; to reneging on policies and regulations; and so much more. All of it, all of it, undermines the willingness and even basic ability of businesses to prudently commit to long-term investments, pursue the advancement of innovative technologies, and train and expand forward-looking workforces—which is exactly what this moment needs.
• Costs of upgrading outdated grid infrastructure. One of the single largest electricity system cost drivers of the past decade-plus has been skyrocketing investments in grid infrastructure. These investments range from catching up on replacing outdated and faltering equipment; to shifting from analog to digital control systems; to boosting the resilience, flexibility, and reach of the system as a whole. While these upgrades are critically important, they’re also expensive—sometimes extremely so, especially when catching up on decades of deferred maintenance. But instead of helping to dampen these costs and facilitate savings, the Trump administration has cut entire offices dedicated to working on these issues, slashed funding and cancelled projects intended to help address these challenges, and increased the costs of hardware and components.
• Costs of preparing for increasing extreme weather events. Extreme weather disasters, from hurricanes to flooding to wildfires and more, have always threatened the electricity grid, but as climate change has increased the frequency and/or severity of so many of these types events, utilities are entering a whole new era in terms of requirements for system preparation and resilience. Beyond increasing the direct costs associated with upgrading electricity grids, the Trump administration is also actively exacerbating the magnitude of the risk these systems will need to prepare for, as the administration unrelentingly hides evidence of climate impacts, denies climate science, impedes climate collaborations, and blocks any and all efforts to limit heat-trapping emissions.

Each of these issues is contributing to higher bills, and because of the actions of the Trump administration, each of these issues threatens to send future bills much, much higher than they otherwise would.

The Trump administration is confronting the electricity affordability crisis with a flame-thrower.

The Trump administration is unequivocally intervening in the electricity sector in a moment of crisis. The problem is, the Trump administration is the crisis, sidelining solutions while supercharging stressors. With policies like that, of course electricity affordability is careening toward a real and true emergency.

Of course, President Trump and his administration haven’t been working alone. In particular, the Republican-controlled Congress has repeatedly enabled, and often outright facilitated, the administration’s destructive agenda, from declining oversight of illegal administration actions, to weaponizing use of the Congressional Review Act, to drafting and passing a budget bill that severely undercut renewable energy while advancing fossil fuels. Alongside, numerous special interest groups have been more than happy to ignore the immediate as well as long-tail consequences of this administration’s actions while seizing on chances to advance their own particular interests.

But through it all, it’s been President Trump and his administration leading the charge.

Right now, people across this country need policymakers committed to implementing real solutions. The Trump administration is doing anything but—and because of that, we’re all paying the price.

I’ve written before about how you or someone you know needs clean backup power (read: not from a fossil fueled generator). The recent winter storm may just be one of those times, or it may be the next event, whether a winter storm, public safety power shut off for wildfire prevention, hurricane, or another emergency. At that time, I also wrote about how electric vehicles can be a solution in times of power outages, acting like a quiet generator with no direct air pollution emissions through vehicle-to-home (V2H) power flow. Short of V2H, the more limited vehicle-to-load (V2L) functionality that many electric vehicles have can also help in a pinch to power a critical devices, such as a refrigerator or fan. V2L is also great for powering tools at a job site or gear on your next camping trip.

In principle, electric vehicles can do more than backup power V2H and off-grid V2L.  Drivers with a properly equipped electric vehicle and home set-up have the option to use their vehicles routinely in coordination with electricity grid needs to lower their electricity bills through bill credits or incentives. Drivers can do this through managed charging (grid-to-vehicle aka V1G) or bidirectional charging (vehicle-to-grid aka V2G).  

V1G and V2G involve charging up vehicles at times when renewable generation is abundant or when demands on the grid are low. In the case of V2G, vehicles also discharge when the grid needs some extra juice. In so doing, V1G and V2G lower costs for the grid by better utilizing grid resources, including cheaper sources of energy, and by deferring or avoiding grid upgrades that would otherwise need to be made to meet the demand for energy on the grid. Lower grid costs mean relatively lower bills over the long term for all electricity customers. UCS analysis has demonstrated the potential magnitude of V1G and V2G grid savings into the future. (spoiler: it’s in the billions per year!)

I’ll come back to V1G and V2G a little later. First, we need to know, what is the market currently delivering in terms of vehicle functionality and what else may you need to use that functionality?  

Bidirectional charging includes a range of capabilities

To do any kind of bidirectional charging, the vehicle must have both the hardware and software to enable that capability. A key component is a battery management system that allows for discharging the battery for a purpose other than running the electric motors that move the vehicle.

Many new electric vehicles can do V2L bidirectional charging. My Hyundai Ioniq 5, for example, has USB outlets in the cabin, and I have the option of using an adapter for the charging port through which a power cord or extension cord can be plugged in. Some Ioniq 5 trim levels also have a regular wall outlet type receptacle under the back seat. My Ioniq 5 is not, unfortunately, equipped for V2H, but many new vehicles are.

Indeed, the good news is that the market is beginning to deliver V2H capability for home backup power in a growing number of electric vehicle models. And drivers are buying those models: V2H-capable models have grown as a share of electric vehicle sales, reaching over 18 percent in the U.S. in 2025. News from Hyundai outlines plans to start full V2H for home backup power with the Ioniq 9, and meanwhile, Hyundai subsidiary Kia rolled out V2H in 2025 for the Kia EV9. Other available models include Chevy’s Equinox, Balzer,  Silverado, and upcoming 2027 Bolt, as well as the Ford F150 Lightning units that made it off the line before production stopped (over 100,000 of them). There’s the Tesla Cybertruck, Volvo E90, and Polestar 3, too. V2H-capable models from Rivian and BMW are anticipated in the near future.

The most sophisticated end of the bidirectional charging capability range V2G. It allows an EV to discharge power to a driver’s house while the house is actively connected with the grid. Grid-coordinated EV discharge may involve offsetting the house’s electricity needs or exporting power through the house and beyond the electricity meter back to the grid.  This is different from V2H that is set up for backup power purposes only, which you may also see referred to as V2H. (It’s confusing, I know!) The main difference is the kind of connection to the electricity grid you pursue with your local utility. There may also be software to update in the vehicle.

Providing consumers with more options for affordable models with V2H is important. Notably the V2H-capable models mentioned above span a range of price points, including the forthcoming 2027 Bolt on the more affordable end (advertised at around $30,000). However, that V2H capability (and V2G capability for that matter) is not universal in new vehicles is a problem. The majority of drivers purchase on the used market, and consequently are subject to the preferences of new car buyers. If new car buyers choose V2H capable models, then more V2H options will be available on the used market. If they don’t, V2H capability on the used market will remain limited, preventing many drivers from having this important option for clean backup power and other bidirectional charging opportunities. That’s where policy can help by ensuring V2H capability (and better yet, V2G) is offered across all electric vehicle models: if all new models have V2H, the more and more vehicles on the used market will have it over time.

So your electric vehicle is configured for bidirectional charging. What else do you need?

For V2L, you may need the charging port adapter. Some models come with it (check the spec sheet posted in the window!), or it may have to be purchased separately.  

V2H, even just for home backup power, takes a bit more external hardware, including a bidirectional capable charger, control system, and other electronics. This is where additional setup cost comes in, though most of the extra cost would be the same as that of a whole house generator or stationary battery system. Currently, most automakers have only OK’ed their vehicles to push power through their own proprietary system or a single, specific third-party provider. Take, for examples, the General Motors V2H kit (also requires their PowerShift Charger) or the Sunrun system for the Ford F-150 Lightning (requires Ford’s Charge Station Pro and a V2H activation fee for some trim levels).

It’s encouraging to see automakers going beyond vehicle capability to ensure their customers have all the components needed to set up V2H at their homes. At the same time, the walled gardens each automaker has created limits driver choice and prevents them from shopping around for the best price and functionality for a V2H rig. This is another instance in which policy can help by requiring interoperability. And I mean implementation of interoperability, not just capability to adhere to the relevant standards.

If V2H isn’t feasible for you, you still have options

When my household first acquired an electric vehicle, I lived in a rented apartment. Even if my Ioniq 5 was equipped with V2H backup capability, I wouldn’t have been able to install the home setup because I didn’t have access or authority to make changes to my electrical panel. Multiunit housing is a tough nut that needs cracking to unlock V2H for back-up power, as well as vehicle-to-grid opportunities for those who would like to participate. There’s much work to be done to develop solutions for V2H in multiunit housing where access to charging can be a challenge in the first place.

Without a full V2H setup, V2L can still be a lifeline in power outages. As I noted before, V2L can power critical devices to help you through an outage. Aside from the ability to plug in a device, an electric vehicle can be a place to escape from very hot or cold weather while the power is out. I contemplated sleeping in the car with the AC running in a post-hurricane power outage in order to cope with the sweltering weather. That’s not something you can do in a gasoline car parked in a garage due to the tailpipe emissions. Luckily for me, that power outage didn’t last long enough for me and my household to have to camp in the car in order to rest at a safe temperature.

Regardless of whether you personally have an electric vehicle, you may still be able to go to a place where one is providing backup power. A growing area of emergency planning is incorporating electric school buses as a source of backup power for schools, shelters, and other community gathering locations. Check out these examples from Illinois and Oklahoma.

Beyond backup: vehicle-grid integration for bargain bills and a better grid for all

V2H for backup will get many drivers thinking about and using their electric vehicles for more than just getting around. That’s a huge mental shift that was once unthinkable for the hundred or so years that fossil fueled vehicles ruled the road. Once you’ve made the shift in thinking, you might be wondering how else you can leverage your vehicle’s capabilities.

That’s where vehicle-grid integration comes in. If you’re set up for V2H backup, the next step to grid-coordinated bidirectional charging. Whether you stick to offsetting some of your energy use from the grid or go all the way to pushing power to the grid, you could earn you some incentives while supporting the operation of the grid. And don’t forget (one directional) managed charging! The UCS analysis mentioned above shows that actively managed charging, which delays or advances charging based on grid conditions, can create significant benefits all on its own.

The availability of grid-coordinated managed charging and V2H/V2G programs depends on what your local electric utility or third party aggregators offer in your area. I’ll cover more on that another time. For now, I’ll assure you that UCS is working with stakeholders in the vehicle-grid integration space to make sure bidirectional capable vehicles and opportunities to use those vehicles are available to more drivers.

It’s like an entire mid-sized city is being built, and most folks I’ve spoken to had no idea it was even in the works.

The proposal raised alarms for a few reasons. A UCS report last fall outlined the way billions of dollars in data center costs are already being shouldered by ratepayers in Pennsylvania and six other neighboring states. Considering my electricity bill has already increased 68% year-over-year, I’m not eager for it to climb even higher. And considering the vast environmental impacts of artificial intelligence use, I’m not eager to see my community pay the physical toll either.

So, I asked Maria Fernanda Chavez, a UCS Energy Analyst and one of the authors of the recent UCS Data Center Power Play report,to help me understand exactly what these proposals mean and what we can do about it.

Q: Let’s start with the source: what—and who—are driving this data center boom?

MARIA CHAVEZ: Data centers are a general term for facilities that house IT infrastructure which stores and processes data for businesses. They have been around for a long time, but in the last few years the heavy investment in artificial intelligence (AI) has driven the development of “hyperscale” data centers that can use as much electricity as entire cities.

These hyperscale data centers—like the one proposed in the outskirts of Philadelphia—are capturing attention nationwide because of just how massive they are, which consequentially means they come with exponential energy demands and more severe environmental impacts, especially if they are powered with fossil fuels. The rapid and staggering degree of investment in AI is driving the desire from investors for these data centers to be deployed as quickly as possible. Big tech companies like Meta, OpenAI, and Amazon are investing billions of dollars into building data centers, but they’re not the only proponents. (My colleague Paul Arbaje has covered one such case in Louisiana, where local utility Entergy fast-tracked three new gas plants for a massive data center, with ratepayers set to subsidize the costs.)

Investor-owned electric utilities also see a benefit in data center development because it means they have more reason to invest in (and earn a guaranteed rate of return on) the energy infrastructure that powers data centers. (And remember, the same for-profit utility companies keen to pass these costs on to you are pulling in billions in profit each year. Wisconsin-based WEC Energy Group just reported pulling in a staggering $1.6 billion in profits in 2025.)

And the fossil fuel industry at large may see data centers as lifelines and jump at the opportunity to power them.

Q: Yikes. If fossil fuel companies are backing these proposals, what does that mean for the environmental impact of these data centers?

MC: Beyond questioning if we should be letting these proposals go through carte blanche, interrogating how we are powering data centers is tremendously important. As it stands right now, without strong clean energy policies or regulatory guardrails, many data centers being proposed would be powered with fossil fuels like methane gas.

Pushing for the construction of new fossil fuel power plants or keeping coal plants alive past their planned retirement age are strategies being pursued by utilities, meaning that a data center popping-up in your hometown might be bringing toxic air pollutants with it. The implications for fetal and early childhood development, cancer risk, and cardiovascular health are dire—and the Trump administration’s recent move to gut common-sense air pollution standards stands to make these impacts more severe.

And that’s just one dimension. A lot of these data centers rely on gas or diesel generators too, which brings harmful air pollution and noise pollution. They also use a huge amount of water, not only to cool the data centers—which generate a ton of heat as part of their data processing—but indirectly from the thermal generators powering these data centers.

Q: And somehow, folks like you and me could end up covering the costs for data centers to be brought online and keep them running?

MC: You’re going to hear me say it again and again, but that’s why ratepayer awareness and ratepayer protections are so critical right now. Folks need to know when these projects are being proposed, and the public deserves a seat at the table when setting the terms for how these data centers operate. Without public input and protections, ratepayers could end up getting stuck with the bill.

The increasing demand in data centers puts more pressure on the power grid, which means increased costs to build more generation, more transmission, and more distribution to keep up. My colleague Mike Jacobs found that ratepayers in seven states, including PA, were on the hook for at least $8 billion over the last two years just for transmission upgrades to connect data centers to the grid. When you consider the fact that, once online, these campuses are going to gobble up immense amounts of energy, it’s easy to see how those costs might end up hitting your wallet and mine.

But, it doesn’t have to be this way. Per our latest report, enacting policies to decarbonize the US power sector will ensure we can meet the projected growth in electricity demand from data centers with clean energy, while avoiding $1.6 trillion from harmful climate and health costs over the next decade. We can also enact comprehensive ratepayer protections so that regardless of the energy policy pathway we take, regular folks don’t bear the burden of data center costs.

Q: Clean energy and ratepayer protections seem like two core tenets worth fighting for.

MC: Certainly—there is no panacea or perfect solution, but these are essential pieces to protect our communities and our environment. In that sense, the implications of this report are twofold. First, communities can use this report to counter industry narratives that toxic fossil fuels are a prerequisite to get data centers online. Clean renewables such as wind and solar combined with energy storage can do the job just as well, without putting our health and climate at risk.

We also lay out six recommendations community members can bring to decision makers to ensure Big Tech companies are paying their fair share and deploying new clean energy sources. The truth is that there is no single decision maker who signs off on a data center.

It ranges from folks in local city councils to state legislators and regulators, to federal agencies and members of Congress. And then there are regulatory decision bodies at the regional level, too. It’s a pretty wide web of folks, and it can feel intimidating to know where to start.

All the same, it is vitally important to understand which decision points occur where. We might see city governments approving permitting that would allow a data center to move forward. At the state level, we might see legislation that sets out the terms for how data centers can operate. A state public utility commission will have a say in how utilities can allocate data center costs. Federally, regulatory commissions can set out terms for covering grid expansion costs.

Q: Wow, that’s a lot of decisionmakers. Where do I even start?

MC: Start by making your voice heard by your state officials—we’ve got this tool set-up so you can contact your representatives and demand they protect ratepayers like you and me. It’s not necessarily a binary “yes or no” on all data centers. We’re asking for common-sense protections that ask the investors pumping millions of dollars into these proposals, and who stand to gain billions in profits, to cough up their fair share of the costs.

And then I’d encourage you to go to the elected officials at the local level, and come prepared with both your concerns and solutions. You can use this data center map to see proposals that are nearby, and share that with your legislator. You can ask for them to advocate on your behalf, and commit to pushing for transparency and accountability.

Q: Transparency feels vital. It’s hard to even know these proposals exist, much less the scale and the implications. Especially for those of us who aren’t scientists!

MC: Absolutely, and that’s why the public deserves information and access to both the stakeholders pushing these proposals and the officials making decisions on these proposals. We have the right to be informed and to be a part of the decision-making and approval process. And we’ve seen communities across the country stand up, fight back, and beat proposals that don’t measure up to these basic standards.

The easiest place to get plugged in is at the local level. Simply put, we must show up at venues where these kinds of proposals are being discussed and made. We need to make sure that communities are informed about what’s happening in their backyards. Ask your neighbors, talk to your friends, show up at city council meetings, show up at public utility commission meetings, and share both your concerns—like health impacts and affordability issues—and what you want: accountability and ratepayer protections.

Your personal voice is so important. Pair the power of your personal story with a clear ask: like accountability for costs, commitments to clean energy sources, reliance on battery storage instead of diesel generators for back-up power, and public disclosures on power needs, water usage and emissions to name a few options. In addition to our report, which lays out these recommendations, there are so many groups who have the expertise and have developed resources to get you started, like this guide from Media Justice and Kairos. You don’t have to start from scratch in your advocacy.

And I’ll close with this reminder: elected officials work for us and should be accountable to us. At every level, they should be putting your best interest first—and it’s important they hear your voice loud and clear.

There is no shortage of reasons to be concerned about this pro-fossil preference—the toxic pollution, the costs of unreliability, and the deadly climate impacts—but I’m going to focus on one of the more technical reasons: conventional power plants are poorly matched to the fast, highly variable electronic loads inside hyperscale data centers. Wind, solar, and batteries, with modern inverters, are a better fit to meet the load growth while keeping the grid stable and reliable.

The grid is changing, and it’s more than just growth

Estimates indicate that load growth from data centers may account for 12% of US electricity use by 2028, and individual data centers can increase a utility’s total demand by 25%. While the scale takes the headlines here, it’s important to note that this is not just more load, but a fundamentally different kind of load.

For over a century, the grid has been dominated by two different types of electric machines: motors and generators. Both are essentially the same device—the only difference is which direction they’re running. On the load side, motors (things like fans, pumps, and compressors) convert electric energy into motion, while on the supply side, generators (like those used in fossil fuel, hydroelectric, and nuclear power plants) convert motion into electric energy.

In fact, if you spin a motor it will generate electricity, and if you run electricity through a generator, it will spin like a motor. The pairing and symmetry between these electric machines on both the supply and demand side has led to an elegantly simple aspect of how the grid stays stable: inertia.

The inertia of electric machines keeps the grid spinning—literally

Both motors and generators have inertia due to the mass of their metal rotors that spin in sync with the grid. In a motor, when the electricity is shut off, inertia keeps the rotor spinning for a few moments, meaning less work is required to get it back up to speed if the power comes back on quickly. On the supply side, generators share inertia like a bicycle built for two (or more): if one rider stops pedaling for a moment, the inertia of the other pedalers keeps providing power to the wheels.

On a grid dominated by motors and generators (especially very large ones), inertia is important because it takes a few moments to change the speed of these machines. Inertia keeps things humming along, absorbing any sudden changes in load or supply while generators adjust their output. This function of inertia is the first component of frequency response, the system which keeps the grid’s frequency at a stable 60 Hz.

Enter nimble inverter-based resources: grid supplies without inertia

When wind, solar, and batteries began to take an increasing share of grid supply, some experts worried about declining inertia. Solar and batteries have no moving parts, and while wind turbines do have rotors, these are almost exclusively connected to the grid through inverters (just like solar and batteries). This enables them to produce more energy, but doesn’t sync their inertia with the grid. Collectively, these are referred to as “inverter-based resources,” or IBRs.

IBRs are built with solid-state power electronics, components that don’t have any moving parts, meaning they don’t have any inertia. While this was initially seen as a liability, it actually means they can do something conventional generators can’t: change their power output nearly instantaneously.

Because conventional generators can’t do this, they rely on inertia to handle sudden changes in power supply or demand on the grid. Then, inertia has to be supplemented with primary frequency response (PFR) to handle larger or sustained deviations in frequency. PFR has historically been provided by mechanical controls—valves and levers—which have their own inertia that must be overcome (and can’t always be counted on).

In contrast, IBRs can respond to grid disturbances instantly and continuously, providing fast frequency response (FFR), which is faster than inertia and long-lasting like PFR. For this reason, some in the industry refer to FFR as “synthetic inertia,” though this is a bit of a misnomer, as there is no kinetic energy involved.

Grid symmetry—what’s old is new again

As IBRs start to take a larger share of generation, the symmetry that once existed between the supply and demand sides of the grid—both previously dominated by electric machines—has begun to fade. The load side has evolved slowly to date, while the generation side is changing dramatically.

But the load side is catching up. In a data center, up to 95% of energy is used to run computing equipment, which is based on the same solid state power electronics as inverters. In fact, the similarities are so close that the Australian grid operator has started referring to data centers as “large inverter-based loads” in its latest system security plan.

Thus with the growth of AI data centers, things are moving toward symmetry again—as inverter-based resources grow rapidly on the supply side, growing demand is also dominated by the inverter-based loads in data centers.

Faster sources are a better match for faster loads

This leads to an obvious question: if the symmetry of electric machines on both sides of the grid once contributed to grid stability, are there benefits to the new symmetry of inverters?

There are several. First, when loads no longer have inertia, energy sources without inertia are better. On the old grid, inertia resisted changes in grid frequency and kept things stable. In the new grid, loads change much faster, and resources that have inertia struggle to keep up.

Second, while data centers’ operations can shift load by hundreds of megawatts (MW) in seconds, new fossil gas power plants can only do 75 MW per minute at best, while older plants are limited to as little as 19 MW per minute. In contrast, batteries and solar can shift their output by 100% in less than a second. Wind output is a bit slower since there’s still some inertia behind the inverter, but all IBRs are more than capable of matching the power swings caused by data centers.

The challenges of faster loads

We’ve known for decades that batteries are really good at responding to rapid changes in power demand, as they’ve long been central to uninterruptible power supply (UPS) systems. Data centers employ UPS systems to instantly transfer power to backup supplies when there are problems on the grid. As one recent data center market overview explains, “Power interruptions over 20 milliseconds can cause data-center IT systems to crash, so UPS transfer time must be faster than this threshold. Batteries are the only widely used technology capable of ramping fast enough” (emphasis added).

But AI presents a new challenge. When a data center is training an AI model, the processors cycle rapidly between performing calculations and sharing results with each other. These cycles are synchronized across all processors in a data center by design, so that each processor has the latest data from all the others for each new set of calculations. During these cycles, processors can swing from 10% of their maximum power demand up to 100% in milliseconds. At the scale of a large data center, this means the total load can swing by dozens of megawatts, equivalent to thousands of homes coming on and off the grid simultaneously.

In addition to the challenge of simply matching this rapidly shifting demand, these massive power fluctuations can also create frequencies which resonate with generator turbine shafts, leading to stress, aging, and even premature failure. The impact isn’t limited to generator rotors and can also affect voltage and frequency regulation, further destabilizing the grid.

Without changes to how the grid is built and operated, these impacts of the growth in data centers create risks to grid reliability and could result in outages or damage to grid equipment and connected loads. If we’re not careful, this could translate to more frequent and longer power outages, affecting everyone who uses the grid.

Are IBRs up for the challenge?

The short answer: yes. A recent research paper from Microsoft, OpenAI, and NVIDIA summarizes the problems caused by this massive, rapid cycling, and recommends three potential solutions. The first two give processors busywork during the low-power phases of the training cycle, essentially wasting energy to reduce the power fluctuations.

The more elegant solution uses energy storage in the form of batteries, an inverter-based resource. Comparing the three solutions, the researchers give energy storage the best rating across six different categories: reliability, performance, energy usage, ability to meet requirements, ease of integration, and lifetime. The other two solutions only out-perform storage in terms of cost. Looking ahead, the researchers note that “for even larger AI training deployments in the future, long storage BESS (battery energy storage system) should also be considered.”

Does it scale?

The bigger question is whether this logic scales—if batteries can address these issues at individual data centers, can we extrapolate the same benefits up to grid scale across IBRs? And if so, can a grid without inertia really be stable?

We’ve already discussed how wind and solar outperform conventional generators when it comes to the speed that data center loads demand, and recent evidence indicates that better integration of IBRs generally can provide the voltage and frequency regulation needed to keep the grid stable better than conventional generators.

The figure below shows that several grids around the world are already integrating large shares of IBRs; large grids like ERCOT in Texas and NEM in Australia can operate stably with instantaneous IBR penetration up to 75%. The article the figure is drawn from concludes: “These examples show that it is possible to operate very large power grids with very high penetration levels of IBRs. By relying on sufficiently many GFM inverters, operation with IBR penetration levels up to 100% is feasible.”

“GFM” refers to grid forming inverters. Whereas traditional grid following inverters (GFL) only work in a grid with synchronous generators (relying on their inertia), grid forming inverters, as the name implies, can create their own grid, and go it alone without inertia.

More points in favor of IBRs

Beyond technical alignment, inverter-based resources offer other advantages during this period of rapid load growth:

IBRs can be built quickly. The figure below illustrates how long it takes different grid projects to be built. Data centers come online in two to three years, which narrows viable supply options in the near term to batteries, solar, and fossil gas, with batteries being the fastest to deploy. And while solar and gas plants can be built on similar timelines, gas faces severe supply bottlenecks that can add up to seven years before construction even begins.

IBRs are cheaper. New wind and solar are cheaper than new fossil gas generation, and batteries are a cost-effective alternative to fossil gas peaking plants, in addition to their role in managing data center load discussed above.

IBRs are cleaner. As our recent analysis shows, data center load growth is expected to be most rapid over the next five to ten years, meaning it’s critical to make the right energy supply choices now. Under current policies which favor fossil fuels, the expected growth in data centers would lead to $1.5 trillion in climate damages and $32 billion in health damages by 2035 due to air pollution and emissions from fossil fuel power plants.

Real power system actors already understand these advantages. Google recently struck a power supply deal to power a new data center in Minnesota with 1,900 MW of IBRs.

Looking forward

For over a century, the grid has been stabilized by spinning metal—large synchronous generators whose inertia helped keep frequency stable. That approach made sense when both sides of the grid were dominated by electric machines. But AI data centers are fundamentally different: inverter-based loads that draw massive amounts of power with rapid fluctuations. As inverter-based resources expand on the supply side and inverter-based loads grow on the demand side, the grid is evolving into something fundamentally different: a system managed less by inertia and more by fast, precise, electronic control.

However, none of this suggests inertia is obsolete. Existing synchronous generators will remain on the grid for decades, and there are still challenges to implementing 100% IBR-based grids that must be worked out in practice. But as inverter-based loads and resources grow, stability will depend increasingly on fast electronic control rather than rotating mass.

The question, then, isn’t just whether wind, solar, and batteries are cleaner or cheaper. It’s whether they are a better technical match for the loads we’re building. Powering 21st-century, inverter-based loads with more 20th-century, mechanically constrained generators risks locking in a mismatch that makes the grid harder to operate. As the grid is evolving, our supply choices should evolve with it.

So what’s coming apart under the stress of all this?

As the information network and data center industry, known for the slogan “move fast and break things” collides with the electric utility world which has slogans 180 degrees opposite, a raft of conventional wisdom is failing. Both the economics and the industry ways of doing things are getting turned upside down. Note at the outset: the idea from State of the Union of customers building electric supply is not new. This was common first with paper and lumber mills burning wood for steam, then in a broad expansion of cogeneration of steam and power at industrials during the 1980s and 1990s, and now with rooftop solar boom of past decade.

Push to increase supply with increased costs

The utility industry originally expanded without raising costs through economies of scale, then with consumer energy efficiency, and most recently with renewable energy. The rush to build supply has scrambled economics with differing emphasis on basic priorities of speed, reliability and cost accountability.

Amongst the data center companies, powerplant economics are a mess. The rush to build supply has bid up the cost of conventional gas-fired plants (and deliveries are back-ordered for 5+ years). To work around that, big tech companies are bringing small truck-mounted turbines designed for short-term or peak use by the dozens. This is the opposite of economies of scale. These generators have roughly 50% higher fuel costs than a utility-grade combined cycle gas-burning plant, and the maintenance of a field of engines is multiples more expensive. That also means 50% more emissions pushed onto the local community, and tempting opportunities for developers to play fast and loose with air permitting rules. This is just a boom-town mentality that jeopardizes everything in the name of speed.

The two dominant existing economic structures for paying for plants reveal serious issues when seen in perspective. Where the old utility structure is still in place and costs are passed on to consumers, utilities are building expensive gas-fired generation and spreading the costs to all consumers. That is, when cost accountability is lost, the costs to consumers aren’t controlled.

In Texas competition is the governing principle. Texas (ERCOT region) does not have protected monopolies in the powerplant business and a costly decision is borne by the owner of that decision. In Texas, wind, solar and batteries are the dominant new supply additions from 2021-2025, and also winning in 2026 on both price and speed to market.  This competitive structure recognizes the falling costs of solar panels and battery storage, and gets new supplies added quickly.

Utilities’ costs claims

Utilities like to say they lower costs by adding more customers, but this old logic worked either when incremental supply was a lower price than previous supply, or there is existing unused capacity to serve new customers—neither are true today. Another utility favorite, making the data centers pay their fare share, breaks down when the data centers evade the basic principle of rates that those who cause the costs must pay the costs. This is most clear with transmission connections, but also with the spreading of costs of new, higher cost power supplies. Much of the proposed legislation seeks to sort this, as business-as-usual policies were established under assumptions that each customer was small and spreading costs was fair and reasonable.

Basic policies upended in current rush

Monopolies

So what’s on the policy-making agenda to allow data centers to transform the electricity industry? The biggie—utilities’ monopoly to connect and sell to new customers is threatened by the most dramatic schemes of data center plans to build off-the-grid, and generate all their own power. This expansion of the concept of a micro-grid has been enabled by legislation in a few states, as it revises the idea of a service territory franchise for the regulated monopoly with an obligation to serve. This scenario requires connections to a gas pipeline, a fiber communications network and probably a water source, but no electric utility connection. The data center developer brings their own electric supply attached to their new demand for that electricity. Utility companies hated this idea when US military bases tried to strengthen their resilience (and war-fighting ability) with power plants on-base. The Defense Department lost that fight in the US but use the on-base, mobile generator-approach when operating off the grid in hostile locations.

Planning

In normal times, utilities’ planning aimed to maintain reliability with gradual change and some direction from state clean energy laws. The data centers’ drive for speed is overwhelming all that. In simple terms, a data center can be built much faster than a new utility power plant. New transmission needs more time too. The largest utility planning organization, PJM, with an area of 13 states, the District of Columbia and the greatest concentration of data centers in the world began changing its transmission planning calculus in 2023 but hasn’t started to use that new approach. In the meantime, PJM has been unable to get new supplies connected to the transmission system fast enough. Late last year, after a few years of record high prices for new supplies, PJM conceded that the supply committed for summer of 2027 is not enough to be reliable as PJM normally defines adequate supply. This is a bad situation.

Expect more wonk in a utility discussion

To fill out this landscape survey, we need a few fundamental assumptions about how the rules work for electric utilities, public health and safety, and competition.

Fed-State jurisdictional boundaries

States have a range of authority related to heath, safety and consumer protection under the US constitution. At the start of his current term, President Trump declared an energy emergency and began down a path of scrambling assumptions and authorities. One extraordinary move has been issuing orders under Federal Power Act Section 202 (c) to keep fossil fuel plants from retiring, regardless of the owners’ economic decisions. These orders also override state air pollution limits, and a lot of other state protections. To speed data center expansion, the US DOE seeks proposed rule changes to widen Federal electric utility regulation with 14 proposed reforms that get into costs, planning, monopolies, obligation to serve described above as largely in the states’ authority. So far, states and utilities have opposed much of this. In a flip-the-script move, Utah requested that the Nuclear Regulatory Commission expand the state’s authority on microreactor licensing, fuel storage and reprocessing.

What new supplies to connect and how

Central to the idea of competition in the US electric power industry is open access to the transmission grid. New plants have been allowed to compete since the 1980’s with the assumption that they would have equal access to getting connected to the grid. After gridlock in that process that’s plagued new competitors, primarily wind, solar and batteries, this assumption is being swept aside. Policy changes, some subtle and others less so, are creating fast lanes and set-asides for new gas generators. Totally lost in all this are the quickest new sources available from virtual power plants linking existing distributed resources and procurements of energy efficiency and capacity. This makes a bad situation worse.

Another key rule for the energy supply broken by the current administration is the validity of permits and leasing processes that have been completed. Offshore wind projects with have been most reported, but federal lease agreements with renewable energy and their permits have been decimated by the Department of Interior. The impact on consumers in the Northeast is clear from this interference, as offshore wind contracted to the states make a significant contribution when gas supplies are exhausted and energy prices are high. In the midst of a bad situation, this is unconscionable.

Looking forward to a new change

In any crisis, new opportunities can arise. At the moment, the assumptions about energy storage (mostly batteries) are evolving quickly. Texas has met considerable demand growth with 2-hour batteries. California has adopted 4-hour batteries. In the data center world, some are saying storage is non-optional. Where utilities and data center companies have seen the future, significant new energy storage is in the plans.

The stress of so many data centers, often incredibly large, pushing demand for electricity is reshaping the electric power industry. The reforms proposed and already made to aid the data centers will have implications for the health and safety, as well as economic well-being of citizens in every state. Both state and federal policymakers have a role. Take this up with them.

This action is monumental in its significance for three reasons.

First, for its direct attack on the nation’s capacity to tackle climate change.

Second, for the pure totemic power of climate denial now adopted as official US climate policy.

And then finally, this: For all its staggering import, the EPA’s new repeal is wildly weak in its execution.

Imagine being so contemptuous of the people across this country staring down the towering costs of climate change that you don’t just attempt to shred the EPA’s capacity to act, but do so with this shoddy, full farce of a bad-faith rulemaking, incoherent in its logic and actively disdainful of facts.

The upshot is a rule that should be taken seriously for the staggering significance of its impact, yes, but not to confuse that with the effort being serious, because on statute, on science, on consideration of impacts: this repeal is anything but.

What is the Endangerment Finding?

The EPA first issued the Endangerment Finding in 2009, making a science-based determination that greenhouse gases, such as carbon dioxide and methane, endanger public health and welfare and that greenhouse gas emissions from vehicles contribute to that endangerment. As a result of that finding, the agency has a statutory obligation to regulate heat-trapping emissions. The agency started with vehicle greenhouse gas emission standards, followed by power plant carbon dioxide standards and oil and gas methane standards.

The EPA did not arbitrarily undertake the Endangerment Finding process: The agency was specifically prompted by the 2007 Supreme Court decision Massachusetts v. EPA, which found that “greenhouse gases” were unambiguously air pollutants under the Clean Air Act. As a result, if the EPA found that greenhouse gases endangered public health and welfare, then the agency would have a statutory obligation to act.

The law was clear, the Supreme Court was clear, and the subsequent science-based finding was unequivocal. Multiple efforts to challenge the Endangerment Finding and the obligations underpinning ensuing regulations were rebuffed.

So, how now does the Trump administration seek to overcome all that? Contortions of law, and avoidance of fact.

What is the Trump EPA’s statutory-based argument for repeal?

In the EPA’s proposed repeal of the Endangerment Finding issued this past summer, the agency put forward a two-pronged argument: first, that the underlying statute, the Clean Air Act, precluded the agency from regulating greenhouse gas emissions, and second, spuriously, that the scientific determination was unsound, with “uncertainties” in climate science undermining claims of endangerment.

In the EPA’s final repeal, the agency only kept the first part, the statutory argument, and claims to have dispensed with challenging climate science. However, as detailed in the next section, the agency in fact still makes numerous arguments that are ultimately entirely reliant on challenging foundational climate science, it just now does so while cloaking them as “statutory” arguments and, in so doing, attempting to bypass otherwise-required evidentiary requirements.

The EPA makes two main arguments in the final rule for its repeal of the Endangerment Finding and vehicle greenhouse gas emissions standards:

• Statutory authority: Trump’s EPA argues, in various ways, that the Clean Air Act does not in fact provide the agency with the authority to issue vehicle greenhouse gas emissions standards, and as a result precludes the agency from even making an Endangerment Finding. The EPA first argues that under the Clean Air Act, “air pollution” must be understood as having local or regional impacts, not global impacts. With this argument, EPA fully side-steps the fact that the Supreme Court explicitly affirmed that greenhouse gases are “air pollutants” under the Clean Air Act, that subsequent cases upheld EPA’s approach, and that—obviously?—pollution with global impacts alsohas impacts felt heavily locally and regionally.
  
  Separately, the EPA is also now arguing that the agency cannot issue a stand-alone finding of endangerment without also considering the impact and efficacy of ensuing regulations—again, fully side-stepping the fact that the courts have repeatedly upheld the agency’s stand-alone finding. But, this new assertion is required for the EPA to then be able to claim that regulating vehicle emissions is “futile” to addressing global climate change concerns (see next point), arguing that it consequently allows the agency to, coincidentally, entirely evade the question of whether greenhouse gases endanger public health and welfare.
• Futility of regulation: Beyond the complete unfoundedness of the “no material impact” argument itself, the EPA also attempts to argue that this claim, which is clearly a question of science, is instead strictly a statutory analysis—a desperate attempt to evade the evidentiary responsibility that would be otherwise required. This clearly doesn’t pass the laugh test, with the agency simultaneously arguing that: first, its rulemaking “does not turn on scientific findings made with respect to the validity, certainty, or extent of global climate change” and therefore does not need to engage with the scientific record, while secondly, also arguing that its non-analysis analysis of eliminating all vehicle greenhouse gas emissions from here through 2100 would result in de minimis climate impacts (without, of course, actually playing those impacts out). How can EPA possibly make this conclusion without engaging with climate science? Of course: it cannot.

These arguments are just staggering contortions of law and fact. This is backed up by the agency heavily leaning the crux of its arguments on two specific Supreme Court decisions:

• West Virginia v. EPA (2022), which established the so-called “major questions” doctrine, finding that it required a clear congressional statement of authorization in “extraordinary” cases when an agency exercises a previously unheralded power, in a manner that is a transformative expansion of its regulatory authority, and economically and politically significant.
• Loper Bright Enterprises v. Raimondo(2024), which overturned “Chevron deference,” where courts deferred to reasonable agency interpretations in instances of statutory ambiguity. Loper Bright instead argued that courts maintain responsibility for determining the best statutory interpretation.

These Supreme Court decisions were worrying at the time of their release specifically for the future threat of how they could be arbitrarily weaponized by the court. Of course, they can also cut the other way—and more critically, it’s misguided to argue that they even apply at all. Time will tell how it plays out for this case, but the very fact that EPA so heavily hangs its arguments on them signals how weak an argument EPA otherwise has.

What is the Trump EPA’s science-based argument for repeal?

When the EPA issued its proposed repeal of the Endangerment Finding this summer, it did so by partially relying on a direct questioning of the climate science underlying the original determination. The EPA’s arguments repeatedly invoked assertions made in a report by the sham-from-the-start “Climate Working Group,” a hand-picked group of five known climate contrarians commissioned by the Department of Energy.

The Climate Working Group report, and EPA’s related arguments, were thoroughly refuted over the course of an overwhelming activation and response by the scientific community, which detailed all the many ways the report misrepresented research, cherry-picked data, ignored peer-reviewed papers, exaggerated uncertainties, and employed deceptive framing. The National Academies also launched a fast-track effort to respond to EPA’s assertions, reviewing work since the 2009 finding and powerfully concluding that the “evidence for current and future harm to human health and welfare created by human-caused greenhouse gases is beyond scientific dispute.”

But that’s not all.

The Climate Working Group was also subject to a lawsuit—in which UCS participated—for violating the Federal Advisory Committee Act (FACA), which is specifically designed to guard against the very thing that happened here, with policy advice arising from a biased, non-transparent, non-public process. And indeed, the court agreed, finding that the Trump administration had violated federal law. 

As a result, it’s unsurprising that in the final rule, the EPA officially distanced itself from the toxic Climate Working Group report and, alongside, any explicit questioning of climate science.

But Trump’s EPA didn’t really let science alone in the final rule. Challenges to climate science are still littered throughout: they’re just now all under the guise of statutory arguments. Theoretically, and not at all coincidentally, this shift would enable the agency to avoid contending with actual science. In fact, though the EPA hinges a significant share of its argument on how climate change works as well as the relative impact of reducing transportation emissions on human health and welfare, the agency entirely—entirely—dismisses the whole of the scientific record in a brief three paragraphs out of a total of 338 pages of public comment responses, stating, “Therefore, comments related to climate science are out of scope of this rulemaking.”

The EPA can’t have it both ways. It’s trying, desperately, but it just can’t reinvent and invert climate science to aid its arguments while also fully declining any true engagement with climate science. The two simply cannot square.

What are the impacts of repealing the Endangerment Finding?

The immediate effect of this repeal is the loss of vehicle greenhouse gas emission standards. The transportation sector represents the largest share of US heat-trapping emissions, in 2024 on their own equivalent to the fifth-largest emitting country globally. Losing vehicle emission standards doesn’t eliminate the full suite of tools available to driving down emissions reductions from the sector, but it is unquestionably the most significant when it comes to vehicle emissions accountability.

There’s more to the repeal of this Endangerment Finding than just the vehicle emissions standards, though.

First, the EPA has relied upon the Endangerment Finding to subsequently issue regulations for other major sources of heat-trapping emissions, including carbon standards for power plants and methane standards for the oil and gas sector. Neither of these are undone by this rulemaking alone, but the Trump EPA has separately begun efforts to attempt their full repeals, and without question rescinding the Endangerment Finding here puts them in an immediately more precarious position.

Second, it’s back to symbolism. The very fact that the nation’s foremost environmental agency, with a mission to protect human health and the environment, is aggressively, recklessly, gleefully turning its back on climate action matters far beyond the bounds of any given regulation.

It matters that the official position of the US government on climate action is inaction. It matters that the official position of the US government on climate science is climate denial. It matters that the official position of the US government on climate accountability is unaccountability.

It matters, all of it. It matters for how people across this country see their government abandoning action as prices spike from climate impacts; it matters for how people around the globe see potential for the world’s progress when the largest historical emitter attempts to well and fully pack it in.

It matters.

It’s sloppy, it’s incoherent, it’s baseless—and it matters.

And that’s why we’re here for the fight.

What comes next?

Not only will this rule be challenged—the legal fight has already begun. And because of both the seriousness of the consequences of this repeal as well as the unseriousness of the pursuit, the Union of Concerned Scientists is proud to be joining in the lawsuit challenging it. 

This lawsuit will take time to play out and may well head to the Supreme Court. In the meantime, everything is cast into uncertainty—for regulators, for the regulated, and for people everywhere. Will this repeal stand? If this rule stands, will a deluge of liability lawsuits follow? Will the power sector and oil and gas standards be next? And on, and on. There is an enormous amount of uncertainty arising from this action.

But for all the uncertainty, two things are abundantly clear.

First, this repeal is a real and true blow to climate action, and a dismal reflection of where this nation’s federal government stands on climate. And there’s every reason to believe this won’t be the last of the Trump administration’s attempted attacks on the nation’s capacity to act on climate.

Second, the science has never been more certain on the present and looming costs of climate impacts, as well as the magnitude of action required to bridge the gap from where we are to where we need to be.

Which leaves us with this: The people are staring down the facts, while the government turns its back. An untenable state, an untenable fate. The only workable path ahead? Doubling down on commitment to action.

Excitement around AI has spurred a proverbial gold rush of data center development in the United States and globally. This rush-to-build is also driving a corresponding growth in electricity demand. A recent UCS analysis found that, without explicit policies to support clean energy investment, this explosion of AI-driven data center deployment will lead to greater health and climate costs, and likely higher electricity costs for homes and businesses. Although this post will not attempt to answer the fundamental questions posed above, my goal is to give the essential background for an informed discussion about artificial intelligence and its uses. With that, let’s dive in.

What is AI?

Before I discuss what AI is in a literal sense, I should clarify what is meant by AI more generally. The figure above is a simplified guide to the expansive field of AI. Broadly, AI can refer to any system that attempts to replicate human intelligence. Under the umbrella of AI is the field of machine learning (ML). The idea of machine learning is to make predictions in new situations by applying statistical algorithms to known data. Recommendation algorithms are a popular example of machine learning, used by companies like Netflix or Spotify. There is a veritable zoo of algorithms that vary according to the kinds of data being used and the questions asked of those data. However, one of the most popular algorithms is called a “neural network”—so-called because it was inspired by the neurons in a brain. The figure below depicts a simple neural network.

In a neural network, some input data activates the nodes in the input layer, which in turn activates the nodes in the hidden layers in a cascading fashion until it reaches the output layer, like the way neurons connect and fire in a brain. As the number of nodes (or neurons) increased with additional computing resources, researchers further categorized larger neural networks under “deep learning.” Interestingly, “deep learning” was the buzzword du jour before OpenAI’s ChatGPT 3.5 was released in late 2022. Now it is AI. This brings us to large language models (LLMs), which are a specific flavor of deep learning architectures. It is in the context of LLMs, such as ChatGPT, Claude, or Gemini, that most readers will be familiar with the term AI. Due to the familiarity with AI via LLMs—and because LLMs are a sub-category of neural networks—the rest of this blog post will focus only on AI in those contexts. 

Now, in a literal sense, AI is comprised of a large file of numbers called “weights” and “biases” that correspond to assumptions. For the sake of this discussion, I’ll stick to calling them assumptions. These assumptions are stored in a computer’s memory. There are many AI models that you can run locally on a desktop computer or laptop. But popular LLMs like Claude, ChatGPT, or Gemini, can take hundreds of gigabytes (GB) of memory to run, which requires numerous powerful computer chips, known as graphical processing units (GPUs).

Storage vs memory, and why AI models run on GPUs

You might be familiar with hard drives or solid-state drives (SSDs) for storing data on your personal computer. The cost of storage has fallen a lot in the past few years such that you can get a 1TB SSD for around $150 at the time of writing. While this could certainly hold an LLM, storage is not where computers make calculations (at least, not quickly). You can think of storage like a storage locker—a place where you can put things you want to keep, but aren’t using regularly.

On the other hand, computers also have something called random access memory (RAM) which is used to store information your computer needs to access quickly. You can think of this like the counter space in your kitchen. This space limits how many pots you can have boiling or how many vegetables you can prepare simultaneously. For your computer, if you have a lot of browser tabs or programs open at once, that will use up a lot of RAM because your computer must keep track of all of those different tasks. For high-end computers with dedicated GPUs, there is a third type of resource called video RAM (VRAM). VRAM is allocated specifically for intense graphics processing for video games, 3D rendering, and now, AI.

How does AI work?

AI works by taking some input or set of inputs and applying the model’s assumptions to that input. Since computers only comprehend numbers, text inputs are first converted into chunks called “tokens.” These tokens are assigned a numerical value which corresponds to an “embedding” or “vector.” These vectors specify where a word lives in the space of all possible words and encode semantic meaning. Similar to every place on Earth having a unique longitude and latitude, every word has a unique vector coordinate.

An AI model’s assumptions are refined through many iterations in a process called “training” where the model is asked a question and then adjusted based on how close it was to the correct answer. For example, we might ask an AI model to fill in the blank: “An ______ a day keeps the doctor away!” At first, the AI model might answer with something nonsensical, like “penguin.” But, because all words are expressed with vectors, we can figure out exactly how far away the model was from the correct answer in its training data, and, more importantly, how the AI model should adjust its assumptions to get closer to the right word.

This training process can take many iterations, and it requires a lot of data to generate good results. This is why AI companies scrape the entire internet to build their training datasets. Once training is complete, the assumptions are fixed, and the model can be used to make predictions when someone inputs a prompt or another piece of input data (this is often called “inference”). AI companies have not published a breakdown for the amount of energy used by training versus inference. Although model training represents a more computationally intensive upfront cost, the share of AI workload devoted to inference is expected to exceed that of training in the near future, if it hasn’t already.

Why do we need data centers to power AI?

So, why do we need data centers to power AI? The answer comes down to scale. GPUs get their speed from the ability to do many calculations in parallel (akin to having multiple checkout lines at a grocery store). However, as mentioned earlier, modern AI models need a tremendous amount of memory.

OpenAI’s GPT 3.5 model has 175 billion assumptions, requiring around 350 GB of memory, and GPT 5 has an estimated 1.7-1.8 trillion assumptions. The most advanced commercially available GPUs (at the time of writing, NVIDIA’s H100 GPU) have 80 GB of VRAM. One instance of GPT 3.5 would require 5 GPUs! For context, a higher-end gaming computer typically has 10 GB of VRAM. With approximately 900 million weekly visitors to ChatGPT as of December 2025, it’s not hard to see why OpenAI, one the major AI companies, owns a reported 1 million GPUs.

Housing these massive collections of GPUs requires specialized facilities ranging in size from enterprise data centers, for smaller companies and organizations, up to hyperscale data centers operated by tech giants. These hyperscale facilities can draw upwards of 100 MW of power, which is enough to power a small city. The electricity demand from data centers exploded by 131% between 2018 and 2023, primarily motivated by AI training and deployment. Future AI-focused facilities are planned at the gigawatt scale.

What does AI do?

What AI does is make predictions based on some input and its built-in assumptions. This means that AI systems do not “know” things. You could ask an AI model “what color is the sky?” and it might say “blue,” but only because its training data included sufficient examples of the sky’s relationship to the color blue. What color is the sky? According to an AI, it is blue (with 98.5223% probability).

AI models also do not learn continuously. As I mentioned earlier, after an AI model has been trained, its assumptions are static. If an LLM appears to “learn” things about you through conversation, it is because that information is still within its context window (the maximum amount of information that an AI can process at one time).

Finally, AI models are not (yet) general intelligence, meaning that models don’t exceed human cognitive abilities across all tasks. Different machine learning models may perform well on some tasks but not on others. For example, LLMs have proven exceptional at generating text and bits of code but are unlikely to be useful for predicting wind speeds or stock prices. This is arguably the most important lesson from this blog post: that all AI models have limitations, and understanding what those limitations are is critical for making decisions about how AI is used.

Where do we go from here?

Whether it’s drafting emails or generating images with LLMs, filling gaps in historical climate records with AI climate models, or media streaming services recommending your next song or TV show, AI is already affecting the way people work and live.

Despite the incredible possibilities for AI, the environmental and social costs associated with its development and use expose serious trade-offs. The fundamental machine learning algorithms that allow climate scientists to create granular models of the Earth’s atmosphere, or allow energy companies to optimize renewable energy generation, are also driving massive growth in power consumption from data centers in the form of generative AI models, like LLMs. The growth in electricity demand from data centers is also driving up electricity costs to consumers. Understanding these trade-offs, and making decisions about them, should not fall to tech companies alone. It requires all of us.

As the UCS analysis discussed at the outset makes clear, the decisions we make now about the way we power data centers and invest in clean energy will shape the costs and benefits of AI for years to come. The more informed the public is about what AI is and how it functions, the better equipped we’ll be to have these conversations.

This scenario is currently playing out in my Maryland city, along with countless local governments across the country. Data centers present many threats to the environment, but perhaps the most troubling is the danger of locking in fossil fuel infrastructure at a crucial moment for arresting climate change’s most harmful effects. There is an alternative: A new UCS study shows data centers can be powered primarily with clean energy like wind and solar, meeting electricity demand while—if paired with ratepayer protections—protecting public health and pocketbooks. This path would also protect the climate by reducing air pollution and heat-trapping emissions from fossil fuels, lessening the chances of stranded assets and environmental damage if the AI gamble doesn’t pan out.

A lack of transparency and accountability around powering data centers is blocking this path, however. Opaque agreements and absence of oversight are enabling industries including oil and gas to inflate demand for their polluting products, increasing risk to the climate, economy and public health.

The decarbonized data center delusion

The largest publicly-owned oil and gas companies are investing aggressively in data centers. These centers’ demand for electricity is not only voracious, it’s immediate: The long lead times required to connect to regional grids and for grid operators to add capacity and infrastructure make the prospect of bespoke energy attractive to impatient tech companies. Thus, many of the agreements between tech and fossil fuel companies involve supplying electricity directly to data centers, bypassing grids and utilities altogether.

In December 2024, ExxonMobil announced the development of a natural gas-fired power plant specifically to supply electricity to data centers—the first time the company has built power plants that don’t serve its own operations. A month later, Chevron announced a partnership with electricity company GE Vernova to develop gas-fired plants that would directly fuel data centers rather than feeding into local energy grids. More recently, ExxonMobil partnered with NextEra Energy to develop a 1.2 gigawatt power plant for hyperscale data centers.

Both companies portray these projects as “low-carbon” because they plan to combine gas with carbon capture and storage (CCS), a process that aims to capture carbon emissions and store them underground. The technology is still unproven at the scale needed to offset the emissions resulting from feeding hungry data centers, however.

That didn’t stop ExxonMobil CEO Darren Woods from hyping its power stations as “decarbonized” in a January 2025 earnings call, where he claimed more than 90% of the emissions from burning gas could be abated. Woods has cited the company’s 2023 acquisition of Denbury Inc., which gave the company extensive CO2 pipeline infrastructure and storage sites along the US Gulf Coast for CCS, as ExxonMobil’s competitive advantage in the data center race.

Not all oil and gas companies are meeting the AI moment with fossil fuels: BP contracted to provide clean power for a Google data center in Indiana via BP’s solar energy subsidiary. Shell partnered with tech companies such as Intel on providing oil-based cooling fluids to lessen the enormous amounts of water data centers are projected to consume to prevent overheating. These examples demonstrate the availability of clean options for data center builders.

AI is not a license to drill

AI’s insatiable need for power is a gift to fossil fuel companies in more ways than one. In addition to expanding oil and gas production, the industry has rushed to leverage AI demand in order to attract investors and financing. For large oil and gas companies, a major vehicle for projecting these demands are in-house outlook reports that paint self-serving pictures of the world’s future energy mix.

BP’s Energy Outlook 2024 addresses AI in its section on the power sector. In both its low-carbon and status-quo scenarios, fast-growing electricity consumption is driven by “rising demand from data centres to support the growing use of artificial intelligence applications,” the report states. Shell names “data centres and AI” among the drivers of increased gas demand in its 2025 LNG Outlook, leading the company to increase its 2040 demand estimates by 60 percent. ExxonMobil’s 2025 Energy Outlook projects global energy demand rising, though it doesn’t explicitly cite AI or related technologies as a driver.

But these projections—based on proprietary methodologies that are consistently higher than independent studies—are based on assumptions that are far from certain. As the UCS analysis notes, all of the major industrial players involved in data centers have an incentive to overestimate demand. In addition to fossil fuel companies, this includes utilities, which earn a guaranteed return on investment, and tech companies locked in a race to attract financing.

Specifically, demand projections are clouded by duplicate proposals from tech companies racing to obtain permits, overbuilding by utilities chasing returns, and uncertainty about the future role of AI in our society, which tech leaders characterize as a foregone conclusion.

State and federal regulators trying to find clarity are challenged by the fact that data center proposals and contracts are typically confidential. This lack of transparency allows industries to hype up demand, increasing danger of an investment—and carbon—bubble. What happens to data centers that don’t get the processing load they were built for? That’s a problem for tech companies. What happens to the climate when oil and gas wells, pipelines, and export infrastructure produce more fossil fuels than data centers need? That’s a problem for everyone on the planet.   

Data centers need transparency and accountability

The data center black box is further darkened by the White House’s efforts to suppress cost-effective investments in wind, solar, and energy efficiency and promote fossil fuels in exchange for industry support. This patronage system extends to the tech sector, with Trump issuing an executive order attempting to prevent states from regulating AI at the same time tech titans rushed to support his $300 million presidential ballroom. By influencing government at the expense of public health and safety, the fossil fuel and tech industries are using a tactic from a time-worn playbook on corporate obstruction of science.

But this heavy-handed selection of winners and losers can backfire. Consumers are already suffering significant surges in energy costs due to AI, and other American industries could suffer as well. Trump has said he’ll make tech companies pay for their energy, but if history is any guide, the government will bail out industries that cause economic crashes, sticking taxpayers with the bill. Meanwhile, the failure to phase out fossil fuels quickly and equitably continues to threaten the earth’s climate and public health, along with political and economic stability.

Transparency is critical for preventing this outcome. Contracts between local lawmakers, data center builders, and power providers should be open to public scrutiny. Fossil fuel companies providing power to data centers must fully disclose impacts to the climate, communities, and the economy. And state public utility commissions need to hold tech companies financially accountable for their electricity use.

Ensuring the electric system is resilient to extreme weather requires similar thinking to how we might ensure our own resilience at home. As someone who grew up on the Gulf Coast of Florida, I think about electric grid resilience similar to how I think about being ready for hurricanes: it requires both near-term preparation when a storm approaches (such as stocking up on fresh water and fueling/charging the car) and long-term planning and investment (such as installing storm windows and maintaining a sound roof). 

Preparing the grid for climate change starts with assessing the risks

When it comes to MISO – the federally-authorized regional transmission organization (RTO) that plans and operates the bulk electric system spanning much of the central United States – it seems focused on the near-term prep for extreme weather, but less so on the long-term planning and grid investments that will reduce the risks of future extreme weather. That’s a significant oversight because right now we are investing tens of billions of dollars into electric grid upgrades meant to serve communities for the next 50 years or more. This leaves a perilous gap in preparing the future grid for the extreme weather and climate changes that scientists are warning us about. This failure leaves communities vulnerable to widespread, long-duration power outages that can have devastating impacts. 

Cost-effectively building a grid that is resilient to climate change and extreme weather begins with two core strategies. First, we need the responsible parties – MISO, states, and utilities – collaborating with communities to establish common goals, priorities, metrics, and strategies for mitigating extreme weather risks. Second, an electricity-sector focused, collaborative, transparent, and science-based risk assessment of climate change and extreme weather must be done. These two strategies can ensure everyone understands the challenges, opportunities, and various responsibilities in achieving a more resilient grid and that investments made at all levels are complementary to each other and protect those most vulnerable. This isn’t currently happening, and it means we’re ill-prepared and potentially mis-investing in our response to climate change even as the sector commits significant investments in building the grid of the future.  

MISO focuses on near-term preparations for extreme weather but leaves the system exposed to long-term risks

Over the course of a year, MISO conducts several exercises with its member utilities and states to prepare for extreme weather events, including:

1. Emergency preparedness exercises where MISO and utilities practice how they will collectively respond to various emergency events including extreme weather. When extreme weather hits, being able to effectively react to power plant or grid failures is critical to minimizing outages and avoiding further damage to the system. Re-routing energy flows, ramping up or turning down power plants, and calling on emergency resources such as demand response must be done with an understanding of how the system and other entities such as neighboring utilities will respond. Successfully navigating extreme weather events takes significant coordination, and these emergency preparedness exercises are central to that effort.
2. Annual system performance assessments: Per federal reliability requirements, MISO also tests the system annually under several “contingencies,” such as the loss of one or more power plants or transmission lines. Which contingencies to study are typically provided by MISO’s member utilities, some of which are in response to historical or perceived threats from extreme weather. During this exercise MISO also tests the system’s ability to move power across the region, a critical capability when extreme weather disables parts of the system. 
3. Seasonal risk and preparedness assessments: MISO and its utility members also conduct seasonal assessments of the MISO system with a particular focus on the summer and winter seasons, when extreme weather is most likely and the system tends to experience its most stressful periods. These exercises focus both on the system–for example, to understand projected energy demand, what resources should be available, and/or which infrastructure assets might be offline–as well as expected weather conditions over the course of the season that will impact energy demand and power plant performance. 
4. Week-, day-ahead, and real-time forecasting and positioning: About two weeks out, when weather forecast accuracy improves, the MISO operations team starts paying close attention to any potential extreme weather on the horizon. Within the one-week ahead window, when load, wind, and solar forecasts become available, MISO develops system operations plans that guide how resources are dispatched to meet demand across the system. These plans feed into the real-time operations where the MISO control room has authority to make severe weather or emergency declarations when extreme weather threatens the immediate reliability of the system. 

All of this is both necessary and prudent, but not sufficient: current practices focus on making sure the current grid survives the current weather-year, but do not look much further than a year into the future and drive little, if any, new investments to mitigate power outage risks during future extreme weather. On the other hand, MISO’s longer-term system planning and investment strategies have that long-term perspective but doesn’t take climate change into account. It’s the latter where MISO, states, and utilities must do better because it will enable the shorter-term exercises and operations to actually succeed as extreme events become more frequent and severe. 

MISO must plan and invest for the world that will exist as changed by climate and extreme weather

There are two primary avenues for new transmission system investments in the MISO system. First are the “bottom-up” transmission projects that are brought to MISO by its member utilities for a variety of purposes including to replace aging equipment, meet day-to-day local reliability needs, or connect new load. These are brought for approval by MISO annually in MISO’s Transmission Expansion Plan (MTEP). 

Then, there are “top-down” projects that come out of MISO’s own planning processes, the most consequential one being its Long-Range Transmission Planning (LRTP) process that has driven much-needed investments in the MISO system. Together, they make up MISO’s multi-year grid investment strategy, and neither give little more than a cursory acknowledgement to the risks posed by climate change and extreme weather: 

MTEP’s contingency analysis:

This is the same annual performance assessment mentioned above describing MISO’s near-term processes with utilities to test the system under various contingencies such as the loss of a power plant or transmission line. While this process, in theory, could help identify grid investments to mitigate the risks of extreme weather, the process suffers from multiple key shortcomings:

1. It is not forward looking: While it is important for MISO to understand how the existing system responds when power plants or transmission facilities unexpectedly go offline, nothing about this process looks ahead to what extreme weather risks might manifest over the coming years or decades.
2. It is disconnected from the realities of extreme weather impacts: This process also fails to recognize when extreme weather causes multiple facility outages at a single time and fails to capture the different probabilities of various facility outages during a range of potential events. For example, certain extreme weather events may have a higher likelihood of causing transmission outages, whereas others may have more direct impacts on power plants. This process fails to consider such dynamics. 
3. It isn’t designed to seek out cost-optimal solutions: Finally, this analysis fails to look at various contingencies holistically in ways that would help identify cost-optimal solutions that might mitigate multiple risks or reduce risks in a least-cost manner. 

LRTP’s “reduced risks from extreme weather” transmission benefit metric

Perhaps the most obvious place where one might expect MISO to do a forward-looking extreme weather risk assessment is during its Long-Range Transmission Planning process. Unfortunately, when MISO evaluates investments needed to meet transmission system needs over the next twenty to forty years, its consideration of extreme weather risks falls woefully short. 

During the LRTP process, MISO evaluates the power resource transition – which power plants might be taken offline or added to the system – as well as expected load growth, policy requirements, and other factors to evaluate how the system will operate and what transmission system investments are necessary to maintain a reliably electricity supply. Potential transmission system investments are evaluated for the benefits they will provide to consumers to demonstrate that the benefits outweigh the costs. One of the benefits included in MISO’s assessment is the benefit of that project to reduce the risk of power outages during extreme weather events. That sounds good. But when you dig into the methodology, you realize that MISO is coming up short: 

1. It is not forward-looking: MISO’s evaluation looks solely at historical events. At no point in the process does MISO ask what extreme weather risks the system might face or how these risks might evolve over the twenty- to forty-year planning period. 
2. It focuses on only one type of extreme weather event: MISO’s process focuses primarily on system conditions during extreme winter weather such as Winter Storm Uri. While Uri was a defining event in the discussion about system resilience, it does not represent the full range of risks the system faces from the various extreme weather events that can impact the system in very different ways. Further, MISO’s approach is focused on conditions where multiple power plants are forced offline but does not consider conditions where transmission facilities may be forced offline either separately or in conjunction with power plant outages.  
3. It assumes everyone experiences a power outage the same: When quantifying a benefit of a project, eventually that benefit is turned into a dollar estimate. In this case, a value is assigned to the cost of a power outage, what is called a “value of lost load” (VOLL). MISO uses one value to represent every residential consumer on the system. (Separate values are assumed for business and industrial ratepayers.) This simplistic approach fails to recognize how communities experience outages differently and how vulnerabilities to outages differ among and within communities. This means the most vulnerable remain so, and it likely perpetuates current inequities in how communities are served by the electric system.

In sum, MISO’s transmissions planning and investment decisions do not adequately consider the long-range risks associated with climate change and extreme weather. A more informed approach would be to incorporate science-driven projections of future extreme weather risks – what events the MISO system is likely to experience, how frequently it should expect such events, and how different communities might experience power outages – over the coming decades to inform its investment strategies. Doing so would better protect communities from long-duration, widespread outages and ensure system investments are providing important resilience benefits to justify the costs being passed on to ratepayers.  

Things to watch in 2026

There are some things to keep an eye on as efforts to address extreme weather risks evolve. In both cases, integrating climate science and community input in a transparent and data-driven manner could begin to unlock the investments needed to keep the lights on in a climate-changed world. 

1. New requirements to plan for extreme temperatures: In June, 2023, the Federal Energy Regulatory Commission (FERC), which regulates regional transmission organizations (RTOs) like MISO, issued new requirements to bolster system planning for extreme heat and extreme cold events. The new rules require transmission system planners such as MISO to develop extreme heat and extreme cold scenarios that can be used to test the transmission system and develop “corrective action plans” to maintain reliability during these events. The details of implementing this new requirement is left largely to the system planners like MISO to determine how robust new planning scenarios are and whether the corrective action plans actually lead to new, resilience-focused investments.  
   
2. Updating planning methods to better reflect the value of investing for grid resilience: MISO’s MTEP and LRTP transmission planning processes are ever-evolving and, to MISO’s credit, are capable of being responsive to system needs and shaped by stakeholder discussions. In 2026, MISO’s processes to incorporate extreme weather risks into its system planning processes will again be up for discussion. This presents an opportunity to improve transparency, incorporate climate science, and be more responsive to community needs in the face of extreme weather risks.   

Next steps towards a more resilient electric grid

Achieving a resilient grid that equitably serves communities starts with understanding the risks facing the grid and ensuring transparency, collaboration, and accountability along the way. To that end, our recently released Power After the Storm report includes these recommendations: 

1. Conduct climate change and extreme weather risks assessments: MISO, states, and utilities need to come together to complete an electricity sector-specific, comprehensive climate risk assessment that looks at the full range of risks facing the grid from climate change and extreme weather. 
2. Engage with communities: Ultimately, the electric grid should serve communities that have differing vulnerabilities to climate change and extreme weather and different priorities in how to build resilience. Engaging with communities and building an inclusive decision-making process ensures grid investments complement other community-centered resilience efforts and will lead to more effective and beneficial investments. 
3. Be accountable to partners: Because achieving a resilient grid involves several actors across the federal, regional, state, and local levels, it will require accountability and each partner doing their part. With respect to MISO, it should be using its deep understanding of the regional electric system, its extensive technical capabilities, and its broad convening powers to bring the various actors together, provide technical support, and ultimately create processes that embed collaboration and accountability throughout. 

Unfortunately, climate change has become a reality, and its impacts will be felt over the coming decades in the form of, among other things, more frequent and intense extreme weather events. We must prepare for this future, including by making smart, science-informed investments in our electric grid that are responsive to the risks and ensure the benefits of a more resilient grid resilience flow to communities in an equitable and transparent manner. As detailed above, there are building blocks in place and clear opportunities to do better, but more is needed if we’re to truly ensure power after the storm. 

But with judges already ruling against the move in each of the five cases—granting temporary injunctions and allowing work to restart—one has to simply wonder, will this administration concede to the inevitable, or will they drum up a new, obscure, job-killing measure to target the industry? 

It can feel baffling, when you grapple with the staggering number of benefits offshore wind delivers, including robust job creation, zero operating emissions, drastic pollution reduction, consequential public health benefits, and critical energy reliability during winter storms. 

So, we turned to UCS Senior Energy Analyst Susan Muller for some insight on the vicious vendetta this administration has against offshore wind (Spoiler: It’s got something to do with the fossil fuel industry and their insatiable thirst for profits.).

Q: What is Big Oil and Gas’ beef against offshore wind, Susan? 

SUSAN MULLER: Fundamentally, they are afraid of losing market share across every power system that has access to the ocean—which is to say, a large part of the country. They’re especially worried about that happening in New England and New York, and to some extent in the mid-Atlantic system. It’s hard to overstate just how much energy is available off the northeast coast, but suffice to say, it’s enough to truly threaten the bottom line for a lot of powerful fossil fuel interests.  

It’s well-established that the wind resource in this area is a world-class energy resource, and the need for energy is concentrated all down the northeast corridor, from Boston to New York to Philadelphia to Baltimore to DC. So, it just makes sense to connect that world-class energy supply off the coast with the energy demand on the coast. We didn’t have that option before the 2010s, because the technology was not commercially mature, but it is very much an option now, and that’s why fossil fuel interests are so scared.  

Q: Why would fossil fuels lose out to offshore wind? 

SUSAN MULLER: The truth is when offshore wind–or onshore wind, or solar resources—compete directly with fossil fuels in the wholesale electricity markets, the fossil fuels will lose every time. These markets are designed to select and run the least expensive power plants, and renewables have a built-in advantage because they run on fuel that is free. So renewables will always be chosen to run, and they will displace power plants that have to pay for their fuel. The incredible UCS graphics team created a great animation that explains how it works.  

Fossil fuel executives, including some former executives who are now government officials, benefit when people don’t understand that this is the way the energy market works. It’s also in their favor when everyone ignores those pesky fuel costs. Those costs—which can spike incredibly high at times—will all be passed on to electric customers. Even worse for family budgets is that the most expensive fuel sets the price of power for the entire market. That means, simplifying slightly, that in a single hour, a $10 per megawatt hour increase in the operating cost of a gas plant can end up being multiplied by tens of thousands, or hundreds of thousands of megawatt hours. So, we can quickly rack up millions of dollars in extra costs which, again, will all be passed on to consumers.

Last year, I was part of a team that showed a larger fleet of offshore wind turbines could have saved households in New England over $400 million dollars last winter, simply by displacing the most expensive oil and gas plants and lowering that market-wide price.  Those enormous savings far outweighed the costs of building the projects, which are recouped over time under long-term contracts.

Q: I’m reading about this freezing cold snap across the United States leading to skyrocketing oil and gas prices and gas wells freezing up. And you’re saying that offshore wind actually excels under these conditions? 

SM: You’ll have to let me get a little wonky here. Freezing temperatures mean more people are burning more gas to keep their homes warm. The result? Gas power plants across the region have less fuel—and because many of them are dependent on what we call “non-firm” contracts, they’re only entitled to the gas that’s left after heating needs are met. 

Then, on the electric side of things, we start to see a mismatch between supply and demand: demand for electricity is above average levels because of the cold (people are using electricity for heat, too), and supply of electricity is below average levels because of those gas contracting arrangements, and—as we learned in Econ 101—high demand and low supply leads to a sharp increase in price.

To make up for the gap in supply, the market turns to older, oil-fired generators that tend to be more prone to outages. The result? Less reliable power. Higher prices for families. I know what you’re thinking: Susan, if only there were an energy source that actually flourished under the blistering cold! 

Well, you feel it every time you step outside in the winter: those bone-chilling, gusty arctic winds. Cold temperatures bring strong winds, and it’s a boon for offshore wind power. 

Q: Aside from savings, that reliability is especially important in the winter because of increased risk of outages?

SM: Yes. Winter power outages are especially dangerous, especially as climate change drives even more extreme weather and a wider swath of the country is dealt with blizzards and storms that the grid simply isn’t prepared for. It’s already a climate-changed world, and it’s time our energy sources and grid reflected it. 

The latest report from UCS, New England’s Offshore Wind Solution, takes a closer look at the weather conditions from last winter to see how the first few offshore wind projects could have impacted grid reliability. The results? Energy from just two projects would have reduced the risk of demand-driven power outages by 55%. Adding generation from two other proposed projects would have reduced that risk by 75%. In other words, our grid would have been much more reliable last winter with those projects operating.

So, not only are we avoiding the toxic emissions that come with burning oil and gas in the winter, we’re looking at a 75% decreased likelihood of a power outage. It’s a public health win, a safety win, a wallet win. It’s appalling to see the federal government come in at the very start of winter and launch spurious attacks, yet again, against offshore wind. 

Q: What is the latest status on the work-stoppage orders issued by President Trump against the five major offshore wind projects in December?

SM: I’m happy to report that the final score is offshore wind, 5 -President Trump, 0.  Judges have ruled in favor of the projects in all five cases, granting temporary injunctions and allowing work to restart. Still, the costs of the delays have been enormous. Workers were pulled off their jobs, specialized vessels sat idle, and weeks of good construction weather were lost. Vineyard Wind was already sending a large amount of power to the New England grid, and that’s helped somewhat in terms of reliability, but we are missing out on the additional power that would have been flowing right now. Revolution Wind was also going to start sending some power in January, and that’s been delayed too, so we’re seeing less reliable power in this cold winter and higher prices for families. The same is true for New York and the mid-Atlantic grid, in terms of delayed power deliveries.

And there’s an even bigger cost in terms of how developers now see the United States.  The President acts unilaterally and blocks private investment into coastal communities, putting thousands of jobs at risk. We’ve got construction crews at sea, teams of workers on rigs trying to get these turbines up and running, and overnight, they’re stranded. What message does that send to investors about the United States as a place worthy of their consideration? The impact is chilling. 

Fortunately, as expected, each and every one of these projects is back on track. And that is just as important an indicator that these projects have momentum, and broad-based support and they will be built, as long as courts continue to uphold the law in the face of these attacks by the administration.

Q: So, if you’ll indulge my bad pun—which way are the winds of change blowing? 

SM: When you talk to the people who are building these projects, local leaders who see the economic boost in their communities, and scientists working in this space, they all have the same message—simply, that offshore wind is worth fighting for. The benefits are just too great to ignore. Even the grid operators, who rarely make public statements, have spoken up to say that they need these projects for reliability.

Countries across the world, from Denmark to China and Australia, are embracing it. And here at home, even with the industry in the federal administration’s crosshairs, support and defense of these projects stretches across ideology, across borders, across party lines. It’s happening at every level of government, from city councilmembers to state reps to governors to the literal Speaker of the House. 

That’s about as big a tent as one can imagine in the current political climate.  

And it’s a demonstration of just how promising offshore wind is as a new power source.  We know that fossil fuel interests will continue to push their false narratives and continue trying to pull whatever strings they can to thwart their competition. But they can only distract from the fundamental value proposition of offshore wind for so long.  Especially as the first big projects come online in the next year or two, more and more people will experience the real-world benefits that they bring. 

And shortly after that, we may have a new federal administration which hopefully will realize that blocking an affordable and reliable power supply for tens of millions of people is just not a smart political move.  

The Louisiana Public Service Commission’s (LPSC) new approach—introduced by Commissioner Coussan, and supported by all other commissioners except Commissioner Lewis—creates a fast track for utility companies seeking to build power plants and other infrastructure in response to a large, energy-intensive customer’s request to connect to the grid. This type of large customer is, by far, most likely to be an AI data center these days.

Substantively, this fight is about blocking the completely avoidable risk of energy-bill spikes for Louisianans, simply to cater to the rushed planning of electricity infrastructure to power data centers—energy guzzling behemoths, which can put city-sized demands on the power grid. There was effectively no way to provide input on the policy without physically showed up to the vote, which a total of three people did. Here’s a look into how this initiative, dubbed the “Lightning Amendment,” became an official, unwritten Louisiana utility policy, with essentially zero public input.

The process, or lack thereof:

Aside from the transcript and minutes of the monthly meeting where the vote took place, this new policy will not be written down anywhere in its final form (for example, through a written order). Transcripts and minutes seem to stay on the LPSC’s website for just two years—but it’s easy to imagine a future in which there are no traces or paper trails for this massive hand-out to tech and utility companies, at consumers’ expense. (So, for the sake of future transparency, we’ve gone ahead and archived the minutes and transcript.)

Let’s talk about the name. The “Amendment” part of “Lightning Amendment” is nominally a change to the Commission’s existing “request for proposals” (RFP) rule, in that it waives those requirements (more on this further down). However, even the name is a farce, as this does not actually appear to be a real “amendment” because the rule in question will not actually be changed in the official, permanent record. The docket where this rule lives shows no such activity.

I’ve spent more time than I’d like tracking regulatory dockets across the country, so trust me when I say this: this is not at all how regulators typically enact new policies, which are generally traceable and permanent so that anyone can find the standards that apply. In the absence of full transparency from the LPSC, here’s our effort to add some permanence and clarity to this new regulatory approach.

The substance: who qualifies for this fast-track review?

So, what criteria would a utility company—say, one seeking to serve a new AI data center—have to meet in order to go through this fast-track regulatory review process— say, for a new gas power plant?

Well, the two main substantive requirements are first, that the utility must have an electricity supply agreement with the data center with a minimum 15-year term, and second, that the data center must commit to covering at least half of the cost of that new power plant.

Yes, just half.

That begs the question: who would pay the other half? In all likelihood, other Louisiana ratepayers, who may not benefit at all from the data center. And the kicker? It could end up being even more than half of the cost funded by everyday Louisianans’ utility bills—more on that below. Keep in mind that these data centers are being built across the country by the likes of Meta Platforms, Facebook’s parent company, and Microsoft, which are some of the wealthiest companies in the world and are entirely capable of paying for 100% of their own power needs.

Ratepayers could fund as much as 75% of capital costs

Electricity infrastructure typically lasts for decades and is paid for by ratepayers until it fully “depreciates.” A depreciation schedule is like a mortgage schedule for paying off the loan taken out for buying a house (there are distinctions between the two, but for the purposes of this piece, we’ll keep it simple).

These terms set up the first trap for ratepayers. A proposed data-center-serving gas power plant, for example, doesn’t fully depreciate over 15 years. Typically this depreciation takes place over a much longer period, around 30 years or more. (Note: a 30-year “depreciable life” would align with Entergy Louisiana’s three gas plants approved last year for Meta Platforms’ data center in Richland Parish.)

In that scenario, we’d have a situation where a utility company could go through this new fast-track process while only securing 25% of the revenue needed to pay for the capital costs of the gas plant: half of the costs for half the depreciable life. Because there is no requirement in the Lightning Amendment for a utility company’s shareholders to share in the costs, ratepayers would be on the hook for the remaining 75%.

In that situation, the LPSC could hold these actors accountable, and protect ratepayers, by  “disallowing” costs from being passed to ratepayers further down the line if the utility acted “imprudently.” A finding of such imprudence would require the utility’s shareholders pay up, instead of ratepayers. However, absent any blatant and visible mismanagement, the Commission is unlikely to conclude the utility acted imprudently after they’ve strongly signaled with this new policy that they’re okay with ratepayers subsidizing Big Tech’s data center development.

And though the language in the minutes is vague, this arrangement appears to be just for the capital costs, meaning the physical infrastructure being constructed. The requirement to cover half of the associated costs does not seem to include operational costs, such as fuel purchases, which Entergy’s customers are also slated to subsidize after the approval of its gas plant project for the Meta data center.

The power costs may also be inflated

Here comes the second trap for ratepayers: eschewing the most basic cost-savings measures. In an effort to speed through these electricity project reviews, the Lightning Amendment waives the Commission’s requirements for utilities to issue a request for proposals (RFP) as long as the utility’s project meets the criteria discussed above.

In theory, the RFP rule requires utility companies to test the market and solicit bids from third-party power providers, such as renewables and storage developers, to ensure that electricity demand is being met at the lowest possible cost. By waiving this rule, utilities do not have to meet the demand of the large customer like a data center at the lowest cost. Instead, they are likely to seek to build their own generating capacity, because it’s often more profitable to do so. This is what happened in the Entergy/Meta case last year: Entergy got a waiver from the RFP rule and elected to build its own gas plants and transmission infrastructure. The utility company now stands to make an estimated $178 million in new annual shareholder profits.

With this new policy, the LSPC is not only setting the stage for half of the costs of powering data centers to be passed to other customers, but also making it quite likely that those overall costs will be needlessly inflated by doing away with the most basic RFP cost savings requirements. It’s hard to envision a bigger gift to Louisiana utility shareholders, at the expense of energy affordability for millions of people.

Planning entire cities’ worth of demand in eight months

From there, the remaining obstacles are small. There are some other miscellaneous requirements, like a letter from the Louisiana Economic Development Secretary “confirming the customer’s interest and the importance of power availability” within five years of an initial application. Once these criteria are met, the Commission is aiming to get the utility’s project approved in just eight months.

In Richland Parish, the Meta data center’s peak load is planned to be twice that of the entire city of New Orleans, and future data centers could also potentially consume entire cities’ worth of demand. Entergy’s gas plant project for the Meta data center was already rushed, originally planned for a 12-month regulatory review process, until it was suddenly shortened to just 10 months. Nearly half a year after the LPSC’s August 2025 approval, Entergy still hasn’t adequately demonstrated that it can power Meta’s massive data center while keeping the grid reliable for other customers.

The attempt to shoehorn this complex regulatory review process—which involves hearings, stakeholder testimony, settlement negotiations, and so on—into a period of just eight months sets the state up for a wide array of heightened risks, namely power outages. Across the country, the rapid growth in data centers is keeping grid-reliability authorities up at night, and Louisianans already experience far more power outages than average. Both transmission grid operators for the state had to implement rolling blackouts last year in two separate events. And though the state’s main transmission grid just managed to stay reliable during the recent Winter Storm Fern, distribution grid disruptions knocked out power for more than 100,000 customers, several of whom tragically lost their lives due to the outage.

The LPSC review process should be aimed at determining whether a given project proposal is in the public interest—something that is going to be much harder to accomplish with a straight face when the “lightning” approach risks even more power outages. By doing away with requirements for assessing cheaper and cleaner resource options, the “Lightning Amendment” is also likely to result in higher costs for ratepayers and more harmful pollution.

Figuring out a better way forward

In the same meeting where Commissioner Coussan’s Lightning Amendment was passed, Commissioner Lewis introduced a very different type of proposal on the same topic. Commissioner Lewis wanted to open a rulemaking docket on handling proposed additions of large loads, such as data centers, which would allow a wide array of stakeholders to provide input on how the LPSC should handle this new challenge of data center growth. However, the vote was deferred in December, and deferred again at the Commission meeting in January.

This proposal is much more reasonable than the Lightning Amendment, and can lead to a framework where ratepayers are protected from the costs and risks of Big Tech’s data center development. If you live in Louisiana—outside of New Orleans, since the New Orleans City Council regulates utility rates there—you can call or email your Public Service Commissioner and urge them to implement data center policy that puts the interests of Louisiana communities above those of Big Tech and utility companies.

Research from Brown University’s Climate and Development Lab, in striking detail, maps a network showing how fossil fuel interests and climate denial groups fund, staff, and strategically guide “grassroots” opposition to offshore wind projects from Massachusetts to Virginia. The result is a facade of local resistance that obscures a coordinated, well-resourced campaign to preserve fossil fuel dependence. (And the threat to fossil fuels is real: a new UCS analysis shows that if a larger offshore wind fleet had been operating, New England’s local renewable resources would have delivered more than double the amount of energy delivered by costly, polluting LNG imports.)

Understanding the battle that’s been brewing off the East Coast is important because, at its heart, it is a battle over who gets to decide our energy future and how disinformation is being used to preserve fossil fuel dominance. Here are some key tactics being used to obscure the discussion around offshore wind.

Manufactured grassroots opposition

Across coastal communities, new organizations have emerged claiming to speak for local residents, fishermen, or wildlife advocates. They show up at town council meetings, flood social media, and file lawsuits with unproven warnings about the “harms” offshore wind could cause.

But when you trace these claims back to their sources, a different picture emerges. The Brown University report Against the Wind documents just how many of these groups are embedded in a dense national network of fossil fuel aligned think tanks, dark money donors, and political advocacy organizations. You can see the full map of these organizations here.

Brown University’s mapped understanding of the connections between individuals and organizations, and the fossil fuel industry, in the offshore wind arena. Source: Against the Wind

These groups share leadership, legal counsel, messaging, and funding streams. What appears to be grassroots activism is, in many cases, astroturf—campaigns that are designed to look like local, community-driven opposition but are actually funded, coordinated, or guided by powerful outside interests.

This is not accidental. As early as 2012, internal strategy documents from anti-wind operatives explicitly called for campaigns that “must appear as a ‘groundswell’ among grass roots.” That guidance has been followed to the letter. Local groups are provided with talking points, media training, legal infrastructure, and scientific-sounding claims that lend an air of legitimacy to opposition efforts.

Fake science as a weapon

Central to this playbook is the strategic misuse of science. Rather than engaging with the substantial scientific record, opponents of offshore wind often lean on unsubstantiated environmental concerns, particularly around whales, to argue against its development.  These claims are repeated endlessly despite overwhelming scientific evidence to the contrary.

Federal agencies, independent scientists, and peer-reviewed research have found no causal link between offshore wind development and recent whale mortalities. In fact, the primary documented threats to North Atlantic right whales remain vessel strikes (and 40% of ships on the oceans globally are actually transporting fossil fuels) and fishing gear entanglement, both long-standing consequences of existing industrial activity in the marine environment.

The persistence of these claims is not about scientific uncertainty. It is about narrative saturation. By flooding local discourse with misleading or outright false assertions, opponents create confusion, delay permitting, and undermine public trust in scientific institutions. This tactic mirrors earlier fossil fuel strategies to cast doubt on climate science itself: if people are confused enough, action stalls.

Intimidating scientists, chilling speech

The consequences extend beyond intentionally creating public confusion and misunderstanding. Scientists working on offshore wind, marine ecology, or climate impacts increasingly face harassment, legal threats, and coordinated attacks on their credibility. Public meetings devolve into hostile interrogations filled with disinformation. Researchers are accused of significant harm, cover-ups, or ideological bias simply for presenting established science.

This intimidation is another well-known tactic in the disinformation playbook. When scientists are harassed or discouraged from participating in public processes, industry-aligned narratives face less resistance. The chilling effect narrows the range of voices shaping public understanding and policy decisions, further tilting the field toward delay.

From tobacco to pharmaceuticals to climate change, industries facing regulation have repeatedly targeted scientists as obstacles to profit. Offshore wind is simply the latest arena.

Bad faith lawsuits as delay

Disinformation alone is rarely enough, so it is often reinforced through litigation designed to slow or derail action.

The same network mapped by the Brown University researchers shows how fossil fuel–aligned organizations provide legal support to local opposition groups to challenge offshore wind projects in federal court. These lawsuits frequently recycle the same debunked claims, but their purpose is not necessarily to win on the merits: it is to delay.

Every lawsuit buys time. Every injunction request stalls construction. Every procedural challenge drains public resources and investor confidence. Meanwhile, while our green energy infrastructure is being delayed, fossil fuel infrastructure, already permitted, already polluting, continues operating without interruption.

This strategy exploits the legal system’s good-faith openness to challenge by weaponizing process against progress at a hefty cost.

Attacking state authority

There is another layer to this strategy that deserves attention. Offshore wind opposition is increasingly tied to broader political efforts to strip states of their authority to regulate energy systems and protect public interests.

Many of the same organizations opposing offshore wind are also involved in campaigns to preempt local climate policies, block building electrification, or weaken environmental review processes.

These efforts are often framed as defenses of property rights or economic freedom, but their effect is to centralize power in ways that favor incumbent fossil fuel industries.

If states cannot set renewable energy targets, enforce environmental safeguards, or pursue climate solutions tailored to their communities, the status quo prevails. Offshore wind, one of the most scalable, near-term tools for decarbonizing the power sector in some parts of the country, becomes collateral damage in a much larger fight over democratic control of energy systems.

The hidden infrastructure behind offshore wind opposition

What makes this all possible is something we don’t talk about nearly enough: disinformation subsidies.

Fossil fuel interests don’t just spread misleading claims: they subsidize the entire ecosystem that allows those claims to circulate, gain unearned credibility, and take root locally. The network mapping that I shared above shows that that local groups opposing offshore wind have received massive and varied information subsidies from national think tanks and industry-aligned organizations including funding, legal support, media content, research reports, expert talking points, and strategic guidance backed by more than $72 million in contributions from just six major donors between 2017 and 2021. These resources allow local groups to project influence far beyond what their size, capacity, or expertise would otherwise permit.

Crucially, this system does not fit neatly into established categories of either purely grassroots opposition or top-down astroturf campaigns. Instead, it operates through dense networks in which local actors remain visibly present and active, while national fossil fuel–aligned organizations shape the narratives, claims, and strategies that local groups deploy. The result is a steady stream of misleading or distorted arguments often framed in the language of environmental concern that are amplified through lawsuits, op-eds, social media, and public testimony.

Scientists, regulators, and communities are facing off with a professionally produced alternative reality backed by coordinated messaging, litigation capacity, and tax-deductible support structures.

A subsidized disinformation system is not a single lie, but a sustained infrastructure designed to delay accountability, demobilize political support, and lock in fossil fuel dependence.

Seeing the disinformation clearly

Offshore wind is a powerful tool for meeting climate targets, improving air quality, reducing reliance on fossil fuels, and a host of other benefits. Delaying it has real costs that are borne disproportionately by low-income communities and communities of color already overexposed to pollution and climate risk, while disinformation exploits and amplifies genuine concerns someone might have about offshore wind.

The opposition movement mapped by Brown University makes clear that these delays are not accidental but a product of deliberate investment, coordination, and narrative manipulation by actors with a vested interest in maintaining fossil fuel dominance. Recognizing disinformation is the first step toward countering it.

Disinformation thrives in the shadows when funding sources are obscured, networks remain invisible, and fake grassroots campaigns go unexamined. Research like Against the Wind helps pull back that curtain, showing how opposition to offshore wind is manufactured, financed, and deployed.

The lesson here extends far beyond wind turbines. As climate solutions scale up, we should expect this industry playbook to be reused again and again. This means that we need to all learn the playbook and confront the systemic disinformation infrastructure designed to keep fossil fuels in power.

The real threat isn’t offshore wind, but a subsidized system of delay that we can no longer afford to ignore.

Given New England’s abundant clean energy resources, it doesn’t have to be that way. New analysis by the Union of Concerned Scientists (UCS) underscores the potential of offshore wind to bolster winter reliability. And the numbers are just the latest strong vote in favor of standing up that offshore wind power in a serious way, and soon.

Winter Storm Fern comes

When Winter Storm Fern swept across the country around the weekend of January 24-25, 2026, it brought with it arctic temperatures. In New England, temperatures dropped below freezing (and, as of this writing, more than a week later, haven’t made it back up in the Boston area).

Under conditions like that, the energy scene in New England follows a familiar pattern, and it stayed true to form during the recent storm:

• The low temperatures drove up demand for heating in homes and businesses. In New England, pipeline gas for furnaces and boilers accounts for the largest share of home heating (40% of households).
• The storm also drove up demand for electricity, including to drive the pumps and fans associated with those furnaces and boilers, as well as to power electric heat (electric baseboards and, increasingly, heat pumps). New England is a heavy user of gas for electricity too, with that fuel accounting for more than half of in-region electricity generation.
• That overreliance on gas put the power sector’s fuel appetite on a collision course with demand from the gas distribution companies that serve New England’s homes and businesses. Because gas companies contract for that gas far ahead of time, while gas power plants look to buy gas only as they need it, when push comes to shove, the power plants lose out.
• The lack of gas pushed grid operator ISO New England (ISO-NE) to “Plan B.” Many of the gas power plants in the region have dual-fuel capabilities, the ability to burn oil instead of gas. Gas is not a clean option for electricity generation, but oil is even dirtier, in terms of both carbon pollution and emissions of pollutants such as particulate matter (i.e. soot) and sulfur dioxide (SO₂). The fuel is dirty enough that, a generation ago, communities fought protracted—but ultimately successful—battles to rein in pollution when oil is burned. To meet the critical need for power during this cold snap, however, ISO-NE sought, and received, waivers on pollution limits from the Trump administration during Winter Storm Fern.

The result of the fallback to an even dirtier fuel, running with no pollution control equipment, was, to quote my colleague Susan Muller, “like an oil spill in the sky”. The lights stayed on, but at a high potential cost to public health.

That chain of events could easily have been much worse this time around. As UCS has detailed elsewhere, a similar winter storm eight years ago led to an emergency declaration by then-Massachusetts Governor Charlie Baker  because of the danger of oil not showing up in sufficient measure. The “Plan B” for reliability was in danger of failing. And “Plan C” in a case like that might feature rolling blackouts, which no one should have to face in such weather.

A much better plan for reliability

A familiar pattern, with familiar—and untenable—consequences. So how do we break free of that pattern?

Experience and data show that more pipelines are not the answer. Even gas-producing regions had trouble keeping gas power plants operating during Fern, as has happened in other winter storms, because of the many ways that electricity generation with gas fails in the cold. New England’s energy portfolio is also already way too concentrated on that one fuel, one that the region doesn’t produce and for which it is at the far, far end of the pipeline.

A better plan is to draw much more on the resources that are already in the region, and that often get more abundant when we need them, not—like gas and oil—less abundant.

That’s where offshore wind comes in. Twenty years of data from ISO-NE show that, in most cold snaps, the same cold-weather systems that strain our grid have delivered large amounts of offshore wind energy to the region at the same time.

That’s also where the new UCS analysis comes in. It focuses on last winter (2024-25) in analyzing offshore wind power’s potential contribution to electricity reliability. It starts with ISO-NE’s own framework for assessing the risk of an “energy shortfall,” when energy supply falls short of energy demand. On a weekly basis in the winter, ISO-NE considers a large set of factors, on both the supply and the demand side, in order to determine when a shortfall is likely.  One important risk factor is “daily energy demand,” which the ISO summarizes with a gauge graphic to illustrate increasing levels of risk. As shown in the graph below, last winter had plenty of days when electricity demand (the black line) was in the yellow elevated-risk zone, and some even in the orange higher-risk levels.

The analysis then uses hourly wind speed data offshore throughout the winter to calculate what wind turbines could have produced.

So how much could offshore wind have helped? The analysis assesses two levels of offshore wind capacity: 1,500 megawatts (MW), which is the total of the two projects under construction/almost completed that will feed power into the New England electric grid under contracts with utilities in Connecticut, Massachusetts, and Rhode Island; and 3,500 MW, which adds in the two other projects selected by Massachusetts and Rhode Island for the next round of contracts.

Even those levels of offshore wind would have made a real difference. The analysis found that 1,500 MW would have cut elevated demand-related risk by 55%, and 3,500 would have reduced it by 75%. That higher level would have kept demand out of the higher-risk zone on all but one day.

That would have meant many fewer days when the old pattern would have been followed, and a lot more days with greater reliability and lower pollution.

Keeping your lights on—without hurting your wallet

Keeping oil burners from firing up matters for more than just for avoiding oil-spill-in-the-sky phenomena, and the reliability challenges that lead to using oil-fired power plants more. Because oil is a really expensive way to generate electricity—even if gas becomes even more expensive at times like those—it also matters for our wallets. A study last year for RENEW Northeast showed that having 3,500 MW of offshore wind online could have saved New England ratepayers $400 million just in Winter 2024-25.

But reliability alone should be a powerful incentive to tear up the dogeared script that has, at times like these, tossed us from one fossil fuel to another. In the lead-up to this past weekend—days after Fern had swept through the region—ISO-NE was again warning of “tight operating conditions,” with peak daily energy demand solidly (and worryingly) in the orange “higher risk” zone, fuel supplies closer to red than green, and gas demand all the way on the red end.

There are all kinds of reasons why the move toward offshore wind makes sense, for New England and far beyond. And why poor policy decisions and still more delays—and the fossil-fueled disinformation that helps fuel those—are a disservice to all electricity users, airbreathers, billpayers, job-doers, and people-lovers.

This won’t be the last winter with extreme cold in New England. If the folks building offshore wind farms off our coast—the pile drivers, electricians, millwrights, pipefitters, ship crews, and so many others—are allowed to do their work, though, this may be the last one that we have to weather without a solid amount of offshore wind at our back.

We can’t change the weather, but we can change how we plan our power system to run smoothly through it all. In New England and beyond, offshore wind can be a central pillar of that plan.

It won’t surprise anyone that solar has been the star of the show, responsible for nearly all of California’s clean energy progress over the past decade. And California’s early investments in solar have helped transform the industry to the point where solar is now one of the least expensive sources of electricity. It’s a remarkable success story.

But there’s an old saying about not putting all your eggs in one basket.

As California’s solar industry flourished, California’s wind industry has languished. Despite the fact that wind is now the number one source of renewable electricity in the United States, California has barely added any wind to its grid over the past decade.

So what happened? Why has California added so little wind to its grid? Is more on the way? And how much more wind does California need in the first place?

How much wind power does California have?

California has a long history with wind power. It was home to one of the first large-scale wind farms, which was built in Altamont Pass in the early 1980s. More recently, the last big buildout of wind power occurred in the early 2010s, with many of those projects built in Tehachapi Pass. Without any big investments since then, California has had roughly six gigawatts (GW) of in-state wind capacity since 2013.

Compared to the rest of the country, something very strange has happened in California. As wind generation increased dramatically nationwide over the past decade, wind generation has only inched up in California. The state’s wind industry has come to a standstill.

But that’s not the whole story. California also imports a big chunk of wind energy from neighboring states. Imports of out-of-state wind have grown slightly more than in-state wind generation over the past decade, but not by much. The net effect is that the total amount of wind generation serving California electricity demand has grown very modestly, by 31%, since 2013. In comparison, solar generation grew by 1000% over that same period. No, that’s not a typo. Wind has just been doing very poorly in comparison to solar.

Why has wind development stalled?

There are a variety of factors that likely led to the slowdown of wind development in California.

For starters, solar got cheap. Renewable energy procurement in California has largely been driven by renewable portfolio standard requirements, which don’t distinguish between solar and wind. So when solar prices dropped sharply, California electricity providers went all-in on solar and procured much less wind.

Another factor is that there are only a handful of good locations for wind power in California, and many of the best spots have already been built out. The best locations for wind also tend to be more remote and therefore require longer transmission lines to connect to the grid, which only increases project cost and complexity. Not to mention the fact that wind projects tend to face more permitting obstacles in the form of local opposition, with eight counties in California enacting significant barriers to wind development.

It’s gotten so tough to develop wind in California that very few developers are even trying to do it anymore. Instead, developers are focusing on solar and storage projects. So even though I’ve heard California electricity providers talk about how they would love to buy more wind power, there just isn’t much to buy.

Does California need more wind?

At this point, it’s reasonable to ask if California even needs more wind. Perhaps the state could just reach its clean energy goals by continuing to go big on solar and storage. And if that were the case, there wouldn’t be much to worry about.

However, all signs point towards a substantial need for more wind power. Various state agencies have conducted analyses to assess the buildout of clean energy resources that will be required for the state to reach its clean energy goals. While the exact mix of in-state, out-of-state, and offshore wind varies significantly between studies, the overall amount of new wind that California needs is relatively consistent. The studies indicate that the state will need 5.2-10.3 GW of new wind by 2030 and 22.5-25.5 GW of new wind by 2045. As a reminder, California currently has a little more than 6 GW of in-state wind, a number that’s hardly budged since 2013. So California would need to kick start wind development for there to be any hope of achieving those buildouts.

All three studies in the table above point towards a big need for more wind power because, as I mentioned earlier, it’s best not to put all your eggs in one basket. The studies all have three key ingredients for decarbonizing California’s grid: solar, wind, and storage. When you take away one of those key ingredients, it’s much harder to achieve the same results. For example, when the California Public Utilities Commission (CPUC) studied what it would take to meet clean energy goals with very little additional wind power available, it found that California would need a lot more solar and storage in addition to much more of a new key ingredient: geothermal.

This illustrates the point that “resource diversity” is one of the keys to achieving clean energy goals. It helps tremendously to have many different technologies (solar, wind, geothermal, hydropower, short-duration storage, long-duration storage, etc.) in many different places (in-state, out-of-state, offshore) to facilitate an affordable and reliable transition to clean electricity. Taking away one of the key technological options only makes it more difficult to maintain grid reliability and more expensive to transition.

Is more wind on the way?

There are three different categories of wind power that California is planning to build: in-state, out-of-state, and offshore. In short, the outlook for in-state wind is not good, the outlook for out-of-state wind is not bad, and the outlook for offshore wind is murky at best.

In-state wind

California’s grid operator, CAISO, manages most of the grid in California. Therefore, most of California’s in-state wind projects go through the CAISO interconnection queue to connect to the grid. Because the interconnection process is backed up and takes so long (though it’s getting better!), a wind project would need to be fairly far along in the process in order to come online anytime soon. So the projects in the queue are what we have to work with for now.

When you look for wind projects in the CAISO’s interconnection queue, there’s nearly 7.7 GW in total, which seems promising! But the more you dig into the details, the more you realize only a small fraction of those projects appear to be viable.

• First, some of those wind projects have already been built, but they’re still lingering in the queue because other components of the project are incomplete. For example, the 265 megawatt (MW, or 0.265 GW) Ocotillo Wind project came online in 2013, but there’s an energy storage component to the project that is seemingly incomplete.
  
• Second, some projects are multi-technology projects that include wind, but the developers seem to have changed their plans and focused on other technologies. For example, the Potentia-Viridi project was ostensibly going to have 400 MW each of wind, solar, and storage; however, a recent permit application for the project only included energy storage.
  
• Third, some projects have been stalled for years and have little, if any, hope of getting built. For example, after being denied permits by local officials, the 200 MW Fountain Wind project entered the California Energy Commission’s (CEC) “opt-in certification program” to bypass the local approval process. However, the Commission also rejected the project late last year due in large part to intense local and tribal opposition.
  
• Finally, some projects are just long shots (or in some cases, may have been speculative all along). For example, Windwalker Offshore would be a 1 GW floating offshore wind project on the central coast of California. But with a federal administration hostile to offshore wind, it’s difficult to imagine this project coming online anytime soon. (More on offshore wind later.)

It’s hard to say with certainty how many projects in the queue are truly viable, but after reviewing the data and asking folks in the industry, I’d say there’s maybe 2 GW that could come online in the next decade. And a handful of projects are clearly happening. For example, the 150 MW Gonzaga Ridge Wind Project is scheduled to come online in 2026, and the 100 MW Keyhole Wind Project is scheduled to come online in 2028.

But collectively, these projects aren’t anywhere close to the amount of wind California has been planning to build.

Out-of-state wind

Out-of-state wind is the only bright spot in the outlook for California wind power. There are three transmission lines that are quite far along in the development process, and those could bring substantial amounts of wind power into California.

• The SunZia Project includes a 3 GW transmission line that will connect to a 3.5 GW wind project in New Mexico. California will receive most of the output from the wind project, at least 2.1 GW, and is scheduled to start delivering energy this year. So California will receive a big bump in wind power very soon!
  
• Transwest Express is also a 3 GW transmission line that will connect to a 3.5 GW wind project in Wyoming. California will have access to at least 1.5 GW of wind on the transmission line. Both the wind project and transmission line are currently under construction, and they should both be operational by 2030.
  
• SWIP-North is a 2 GW transmission line that will bring Idaho wind power to California, though California will only get 1.1 GW of that capacity. Construction is to begin this year, and the line is scheduled to be operational by 2028. However, the federal government threw a wrench in the mix with its cancellation of the 1 GW Lava Ridge Wind Project, an Idaho wind project that could have sent power over the line to California. While development of the transmission line is still going ahead, the energy resources being built at the end are still in flux.

Collectively, these transmission lines could bring many gigawatts of wind power to California. Assuming all goes well with these projects, California is roughly on track to reach its out-of-state wind targets in 2030.

But longer term, California will need even more out-of-state wind, and beyond these three lines, I’m unaware of any other projects in the works. That’s a problem because these transmission lines have taken decades to build, with the process beginning in the mid-2000s (if not earlier). Since multi-state transmission projects take so long, that means it may be quite a while before California has more opportunities to increase its supply of out-of-state wind.

Offshore wind

If you’ve been following the news, you may be aware that the current federal administration is not particularly fond of offshore wind. But despite construction pauses and funding cancellations, California is still going ahead with its big plans for offshore wind.

In 2024, shortly after the CEC established a goal to build 25 GW of offshore wind by 2045, the CPUC ordered the procurement of up to 7.6 GW of offshore wind to come online by 2037. The state will start trying to procure those resources next year, which would represent the first big tranche of offshore wind development. In the meantime, the state has set aside hundreds of millions of dollars for offshore wind port development, which is now starting to be spent.

But building up the offshore wind industry isn’t going to be easy. To build offshore wind in federal waters, developers may decide to wait for a federal administration that isn’t so hostile to the industry. I think that, at best, California offshore wind will be delayed.

What should California do?

California is going to need resource diversity to achieve its clean energy goals. Solar and storage will get the state a long way toward its goals, but it will need other clean energy technologies as well. Technically, California could reach its goals without much more wind if another clean energy technology came along that was cost-effective and scalable, for example, next-generation geothermal. But barring any dramatic technological advances, California is going to need a lot more wind.

I think it’s clear that California’s goal of adding 5.2-10.3 GW of new wind by 2030 and 22.5-25.5 GW of new wind by 2045 will be difficult to attain at the rate things are going. The first step will be to get early development activity going again both for in-state and out-of-state wind projects beyond the current pipeline. To do that, the state may need to identify and reduce barriers to in-state wind development and initiate the process to build more transmission lines that enable both in-state and out-of-state wind projects. At the very least, the state should study what it will take to meet clean energy goals without so much wind power to assess the alternatives, which is already in the works.

There likely isn’t going to be one single solution, but the first step is to acknowledge the problem: California is falling behind on wind development. The second step: let’s fix it!

Americans set few everyday expectations for science, but they are fundamental: We expect the weather forecast to be right, we expect science and technology that allow weather hazards to be anticipated within reason, and we expect public services to protect our lives and livelihoods from such hazards—floods and fires, tornadoes, and hurricanes.  

Well, the fulfillment of those expectations is in real doubt now that the Trump administration plans to dismantle the National Science Foundation’s (NSF) National Center for Atmospheric Research (NCAR), a federally funded institution that underpins critical science that Americans rely on. Administration officials have argued that NCAR’s work can simply be redistributed to other institutions without loss. But NCAR is not just another research center. It is a purpose-built critical infrastructure, designed to integrate observations, modeling, supercomputing, and applied research in ways that no single university, agency, or contractor can replicate on its own. 

Although Congress rejected the administration’s proposed funding cuts to NSF, the most recent spending bill did not include explicit language protecting NCAR as a unified entity.  

As a result, the center remains vulnerable—not through outright defunding, but through fragmentation. The administration could try to cut interagency contracts that NCAR relies on to fund its staff, lay off staff, and relocate critical capabilities. NSF has already outlined plans to restructure NCAR, including moving its supercomputer to another site and transferring or divesting research aircraft it operates. These risks would hollow out the institution itself, breaking apart integrated teams, disrupting continuity in projects, and weakening the unique collaborative model at NCAR that accelerates scientific progress in weather, water, climate, and space weather.  

This distinction matters. NCAR’s value does not lie solely in the science it produces, but in how that science is organized, sustained, and shared across the nation.  

Here are five of the many ways Americans will lose the benefits of scientific research if plans to dismantle NCAR unfold, and two ways we can work to prevent it.  

1. Air travelers will lose protection

Every day, millions of Americans board airplanes expecting to arrive safely at their destinations. What most passengers never see is the science working behind the scenes to keep flights safe through better understanding of atmospheric conditions such as turbulence and microburst winds. 

Turbulence alone is the leading cause of injuries on US commercial flights and cargo operations, and NCAR research has played a central role in reducing that risk by improving how turbulence is detected, predicted, and avoided. NCAR scientists helped develop advanced forecasting techniques that allow pilots and dispatchers to reroute aircraft away from dangerous air currents before passengers are ever put at risk.  

In addition to safety, NCAR research has reduced the $100 million financial strain severe turbulence costs the US aviation system every year through aircraft damage, inspections, medical costs, and delays.  

NCAR’s contributions to aviation safety extend well beyond turbulence. In the 1970s and 1980s, NCAR scientists led research that identified and explained microbursts, a poorly understood weather phenomenon consisting of powerful downdraft winds produced by thunderstorms. Microbursts had caused multiple fatal airline crashes during takeoff and landing, and NCAR findings convinced the Federal Aviation Administration (FAA) and international aviation authorities to develop radar warning systems to detect these threats. Since these tools have been deployed, fatal US airline crashes caused by microbursts have effectively been eliminated. 

Dismantling NCAR and moving this work elsewhere would break the integrated system that makes aviation safety research effective in the first place. NCAR uniquely brings together long-term observational data, advanced modeling, specialized instrumentation, and direct operational partnerships with agencies like the FAA under one roof. Fragmenting that capacity across multiple institutions would disrupt decades of trusted, public-service relationships with the aviation community, making it harder and slower to translate research into real-world protections for pilots and passengers. With millions of people in the sky every day, this is not a risk we should take. 

2. Food security and the US agricultural economy will be at risk

Agriculture contributes hundreds of billions of dollars annually to the US economy, and food security remains a national priority, making NCAR’s research crucial to this weather-sensitive sector. Drought, heat waves, and floods are recurring stresses that impact what crops farmers can grow, as well as food prices for consumers.  

NCAR research is directly relevant to food security. For example, NCAR scientists are working in conjunction with universities in Kansas and Nebraska, and the US Department of Agriculture, to develop CropSmart, a next-generation system that aggregates weather forecasts, crop data, soil conditions, and other inputs into actionable, decision-ready information for farmers, agribusinesses, and agricultural officials. Early projections from CropSmart suggest that if advanced decision support systems like this were adopted on even half of irrigated farms in a state like Nebraska, farmers could save up to 1 billion cubic meters of water and $100 million in irrigation energy costs annually while also cutting about a million tons of greenhouse gas emissions per year. 

If NCAR is broken up, we lose this economic opportunity and the myriad ways it supports US agriculture. NCAR’s long-standing collaborations, integrated modeling and computing capacity, and role as a trusted public-service institution are what allow farmers to rely on consistent, decision-ready information year after year.  

All the agricultural tools housed, supported, or innovated by NCAR would be put at risk, leaving farmers with fewer early warnings, less reliable guidance, and greater exposure to weather extremes. These losses would translate to the food on our tables having a higher price tag, which inevitably increases food insecurity, already a significant problem in the United States.  

3. US national security and military readiness will be weakened

The US military depends on weather and climate intelligence to operate safely, effectively, and strategically. From flight operations and naval deployments to training exercises and base infrastructure, weather conditions shape nearly every aspect of defense readiness. When forecasts are wrong or incomplete, missions can be delayed, equipment can be damaged, and personnel and our national defense are put at risk. 

NCAR’s research and operational tools provide the environmental intelligence that defense planners, operators, and test authorities rely on to keep us safe. Accurate, NCAR-enhanced forecasts have saved the US Army millions of dollars by reducing weather-related test cancellations and avoiding needless mobilization costs. NCAR weather forecasting tools have been used for defense-related purposes including anti-terrorism support at the Olympic games, protection of the Pentagon, support for firefighters, and analysis of exposure of our military personnel to toxins. 

The strategic value of this work is reflected in the breadth of defense agencies that rely on NCAR today. NCAR maintains active partnerships and contracts with the Air Force, the Army Corps of Engineers, the National Ground Intelligence Center, the Defense Threat Reduction Agency, and the Army Test and Evaluation Command. These relationships exist for a simple reason: Accurate environmental intelligence reduces risk, lowers costs, and strengthens national security. 

Dismantling NCAR is a national security threat. Defense agencies rely on specialized, mission-critical environmental products and expertise that are developed, maintained, and refined through streamlined, long-standing relationships with NCAR scientists. These capabilities cannot be replaced quickly without disruption, and even short gaps in trusted weather and environmental intelligence would increase operational risk for current and future missions. Protecting NCAR is an investment in military readiness, operational efficiency, and the safety of those who serve. 

4. Americans in disaster-prone areas will have less time to prepare for, and evacuate from, extreme weather

Since 1980, weather hazards have cost the United States thousands of lives and more than $3.1 trillion. In 2025 alone, disasters cost nearly 300 lives, and $115 billion in damages to homes and businesses. And these weather hazards are expected to worsen because of our changing climate.  

A 2010 National Academies of Sciences, Engineering, and Medicine study found that public weather forecasts and warnings deliver roughly $31.5 billion in annual economic benefits in the United States. These gains in preparedness and economic benefit would not have been possible without sustained scientific research from NCAR. 

Hurricane forecasting provides a clear example of how NCAR research has secured the safety and mitigated the economic losses to residents and businesses. Since 1980, hurricanes have caused nearly $3 trillion in damages in the United States.  

For decades, NCAR scientists have worked to develop and refine instruments and methods to collect real-time hurricane observations and improve our understanding of storm behavior. By the 1980s, data and modeling advances emerging from NCAR research were being used operationally by the National Oceanic and Atmospheric Administration (NOAA), contributing to a roughly 20%–30% improvement in the accuracy of hurricane track forecasts compared to earlier decades.  

NCAR continues to enhance forecasting capabilities for hurricanes, and as well as their associated flood risks through the center’s sophisticated flood risk model. Today, the model is used operationally by the National Weather Service in more than 3,800 locations serving 3 million people. 

If NCAR’s role in advancing forecast science is weakened by dismantling, these gains in disaster preparedness will be put in jeopardy. Forecast improvements do not happen automatically; they require sustained research, coordination, and testing. If NCAR’s research capabilities to develop and improve weather forecasting disappear, the United States will face a major public safety risk. 

5. Americans lose a unique source of national pride

NCAR was never designed to serve a select few. It was built with public investment to serve the nation as a whole. From its founding, NCAR embraced the idea that understanding the Earth system—its atmosphere, oceans, land, and ice—requires collaboration across institutions, disciplines, and generations, not isolated efforts working in parallel.  

That collaborative model is embedded in how NCAR operates. It is stewarded by a consortium of more than 120 colleges and universities across the United States, representing a wide range of regions, institutional types, and scientific strengths. This structure allows knowledge, tools, and expertise to flow across the country, connecting large research universities with smaller institutions, federal agencies with academic scientists, and fundamental research with real-world applications for the public and private sectors. The result is a shared national capability that no single institution could sustain on its own. 

There is something deeply American in that collaborative vision, a belief that publicly funded science should be openly shared, collectively advanced, and used to strengthen the common good. NCAR represents what is possible when a nation chooses to invest in science as a public good. 

For more than six decades, NCAR has shown that open, collaborative science can save lives, support economic resilience and national defense, and expand opportunity across generations. Preserving and celebrating NCAR is choosing a future where shared knowledge, innovation, and public-serving science continue to thrive. 

What we must do now

This moment demands more than concern, it requires action.  

First, NSF is requesting feedback from the public regarding its intent to restructure NCAR. Feedback “will be used to inform NSF’s future actions with respect to the components of NCAR and to ensure the products, services, and tools provided in the future align with the needs and expectations of stakeholders to the extent practicable.”  

Respond, and inform NSF on the value and benefits of all of NCAR, not only its constituent parts. Readers can submit comments by March 13. 

Second, Congress ultimately holds the authority to fund and protect NCAR, and lawmakers need to hear clearly that dismantling NCAR would put the health, safety, and financial stability of Americans at risk. By October 2026, Congress will address the funding of NSF for next year; we must actively and consistently reach out to our congressional representatives now and throughout the year.  

Readers can contact their members of Congress through easy-to-use resources provided by AGU and the Union of Concerned Scientists.

To examine the effects the AI boom is bringing to the power grid, the Union of Concerned Scientists conducted an analysis published today entitled Data Center Power Play: How Clean Energy Can Meet Rising Electricity Demand While Delivering Climate and Health Benefits. In addition to a national report, the project also includes state-specific analyses for Illinois, Michigan, and Wisconsin.

We examined how different levels of electricity demand growth, driven largely by new data centers, affect the power system in the coming years—and how improved policies can respond to these changes with clean energy instead of fossil fuels and protect people from higher utility bills and public health costs.

UCS’ director of energy research and analysis, Steve Clemmer, writes about our national findings and recommendations here. In this blog post, we’ll zoom in on our trio of state analyses in the Midwest to compare what we found, where similarities and difference arise, and what conclusions we can draw.

Illinois: Avoiding increased fossil emissions and spikes in power imports

Our look at Illinois demonstrates the urgent need for action by state policymakers to ensure that rising data center electricity demand is met with clean energy. Illinois does have existing, strong power sector decarbonization policies, most recently enacted through the 2026 Clean and Reliable Grid Affordability Act which built upon 2021’s Climate and Equitable Jobs Act. Yet, our results show the shifting landscape caused by power-hungry data centers requires further action from policymakers.

Using a projection of future data center additions based on various industry sources, we found that data centers will account for up to 72 percent of Illinois’ electricity demand growth by 2030 and up to 65 percent by 2035. Without additional policy protections, over the next 25 years, this type of data center-driven load increase could put Illinoisans at financial risk for up to $37 billion in additional electricity system costs.

Moreover, absent stronger policies such as clean energy requirements for data centers, the load growth will also lead to increased pollution from Illinois’ fossil fuel power plants and rapidly escalate the state’s reliance on out-of-state electricity supply (see figure 1). Both trends are at odds with the state’s policy vision of phasing out gas and coal plant emissions and replacing fossil power primarily with wind, solar, and battery storage that directly benefit Illinoisans.

Conversely, our Illinois analysis shows that more robust policy scenarios at the state and federal levels could address data center-driven demand growth, resulting in savings of up to $2.8 billion in health costs and $112 billion in climate damages globally between 2026 and 2050.

Stronger policies stimulate more clean energy development in Illinois, which means less reliance on fossil plants and less need to import power from elsewhere. It also means keeping more jobs, tax revenue, and economic activity that comes from building wind, solar, and energy storage in the state.

Illinois does not have to accept more pollution and less clean energy development from data center demand growth. To mitigate the worst impacts of data centers shown in the UCS analysis—and to pursue a future more like our improved policies scenarios—we provide a series of recommendations focused on data centers. Most notably, legislators should mandate that they secure new sources of carbon-free electricity and should protect other customers by requiring data centers to pay for additional grid infrastructure and operating costs they cause.

Michigan: Closing the fossil power export loophole

Like Illinois, Michigan has also enacted significant clean energy legislation through a series of bills in 2023. But, just like in Illinois, further policy action is needed to protect Michiganders from the risks that data centers derail progress on climate goals, raise costs, and weaken the grid.

Based on our assumption of the size of data center build-out, our Michigan analysis found that data centers will account for up to 57 percent of demand growth by 2030 in the state, resulting in cost increases up to $51 billion and up to 25 million tons of increased heat-trapping emissions (nearly equivalent to emissions from Michigan’s entire industrial sector in 2021).

A key consequence of the increased demand is that combustion of fossil fuels to generate power that is sold outside of Michigan will increase.  As enacted, Michigan’s clean energy laws only apply to electricity sales within the state, and Michigan typically exports only minimal energy. However, our analysis shows that with the growth in demand from data centers, this will change; Michigan utilities will export more than 56 gigawatt-hours of electric energy to other states annually by 2050, causing CO₂ emissions to increase despite the clean energy laws.

To address this, we recommend a CO₂ reduction policy that would apply to all electricity generation produced in or imported to Michigan, closing the loophole that allows utilities to keep burning fossil fuels. This policy would drive Michigan’s electric sector to net zero emissions by 2050 with an energy mix dominated by wind and solar (see Figure 2). This in turn results in $3.5 billion in reduced health costs locally and $408 billion in reduced climate damages globally by 2050.

In concert with clean energy policies, we recommend measures similar to our Illinois recommendations to require transparency from utilities and their data center customers, create incentives for continued clean energy development, and prioritize the needs of Michiganders by ensuring data centers pay their fair share while also boosting funding of consumer energy programs such as bill assistance and energy efficiency.

Wisconsin: Forging a healthier future through net-zero emissions

The third state featured in our Data Center Power Play analyses, Wisconsin, does not yet have comprehensive clean energy policies in place. To explore the benefits of doing so, we modeled a scenario where the state adopts a Clean Energy Standard and CO₂ reduction policy. Together, these policies set Wisconsin on a path to net zero power sector emissions by 2050, a goal already in place in numerous states.

The data center development boom in Wisconsin further complicates the state’s ability to plan for long-term affordable energy or move away from costly fossil fuel overreliance. Our analysis found that data centers could be responsible for 26% to 36% of the total electricity system costs through 2050. To add to that, under current policies, data center load prompts large additions of fossil gas capacity, while decreasing wind capacity. This results in 31 million metric tons of carbon emissions accumulating by 2035 and 140 million by 2050 as a result of new data center load (see Figure 3).

Implementing and committing to clean energy policies now could drive the state toward a more sustainable alternative while avoiding $8.7 billion in health costs, preventing $428.2 in climate damages, and saving lives in the process. Data centers must bring their own clean energy to power their facilities. To further ensure that these benefits are realized for the communities in Wisconsin that need them most, UCS also recommends that—regardless of the energy policy pathways taken—the state adopt strong ratepayer protections. Consumers must not be left on the hook for the costs that data centers incur. Like in Illinois and Michigan, transparency over data center impacts is key to the plans enumerated above. Decision-making with respect to data center development can’t take place without equipping the public with information and meaningful engagement with communities.

Importantly, Wisconsin must introduce Integrated Resource Planning (IRP) requirements that provide a long-term energy resource roadmap. This keeps utilities accountable for their grid investments, empowering Wisconsin to look ahead towards a cleaner and healthier future in the direction of its state motto: Forward.

There is a better path forward

Looking collectively at our findings for Illinois, Michigan, and Wisconsin, it’s clear that strong, foundational state clean energy policies are helpful for confronting the large—yet highly uncertain—data center-driven growth in electricity demand. But without careful, specific attention by state policymakers and regulators to data centers, the rapid rise in the need for power leads to increased costs and pollution.

The good news from our Data Center Power Play analysis is that there is a better path forward. Improved policies at the state and federal levels can help meet data center load growth with clean energy while protecting ratepayers, reducing health costs, and cutting climate damages.

Special thanks to UCS analysts Lee Shaver and Maria Chavez for their contributions to this blog post.

Electricity demand is increasing, and so are electricity costs for households and businesses across the United States. After more than two decades of relatively flat electricity demand growth, the near-term surge in new demand from “hyperscale” AI data centers is poised to send electricity prices through the roof without strong ratepayer protections and other policy interventions.

Addressing these new threats from data centers—alongside existing price drivers such as increasing fossil gas prices and the high upfront cost of necessary infrastructure upgrades in the face of climate-fueled extreme weather—will require concerted efforts to address both electricity demand and supply. Unfortunately, actions taken during the first year of the Trump administration are moving us in the wrong direction and could result in much greater increases in electricity costs, while also risking more outages, more pollution, and more climate-damaging emissions. A new analysis from the Union of Concerned Scientists (UCS), Data Center Power Play, makes clear just how much is at stake, finding that:

• Data center demand is highly uncertain, but will be the greatest near-term driver of load growth
• Current policies risk putting the United States on a path to meet much of this demand with fossil fuels
• Stronger policies can shift our path to running primarily on renewables without sacrificing reliability, while also bringing important health and climate benefits
• The looming data center-driven demand surge comes with significant costs across all scenarios, and policymakers must take steps to ensure that these costs are not passed on to consumers

Overall, our modeling demonstrates that clean and renewable energy can meet the challenge of load growth from data centers, but policymakers must be proactive to protect our health, environment, and financial interests. 

Trump admin sidelines clean energy amid rising electricity demand

Even as the Trump administration boasts of leading the global AI race, it is sidelining the clean, plentiful, affordable, and efficient energy technologies that are best poised to power such a future.

Even with investments in energy efficiency—which the Trump administration is aggressively undermining—we will still need to significantly increase electricity generation to meet surging demand growth from data centers and electrification. This will require building as much new, low-cost capacity as quickly as possible, and making it as clean as possible to avoid negative climate, health, and environmental impacts. Here, renewable energy technologies, especially solar and wind combined with energy storage, check all the boxes. They have a distinct advantage over fossil fuels and low carbon technologies such as new nuclear and carbon capture and storage (CCS), which have long lead times, high costs, and are largely unproven technologies.

Combining wind and solar with battery storage, while requiring greater data center flexibility, offers a promising solution for powering data centers and meeting overall electricity demand. Over the past few years, the vast majority of new capacity additions in the United States, and globally, have been wind, solar, and battery storage. These technologies also dominate new capacity waiting to be developed in transmission interconnection queues.

Further, the costs for wind and solar have fallen 70 to 90% over the last 15 years, making them cheaper in many cases than new gas and existing coal plants, even without subsidies and before factoring in environmental and health benefits. The lead time to plan and build wind and solar projects is also much shorter than it is for new gas or nuclear plants. In addition, recent studies have identified demand-side flexibility at data centers as an important strategy for reducing peak demand and lowering costs.

In striking contrast to this forward momentum is the Trump administration’s stated preference for using dirty, outdated fossil fuel generation to power emerging technology. At the same time, it has created significant headwinds for wind, solar, and storage, including rolling back federal tax credits, illegally rescinding clean energy grants and financing approved by Congress and under contract, illegally halting offshore wind and other renewable energy projects that are either under construction or fully permitted after lengthy reviews, blocking new wind and solar permitting from moving forward, and more.

The stakes are high. The path that states and the nation choose on how to power data centers has wide-ranging implications for energy affordability, reliability, public health, the climate, and the economy as a whole.

Powering data centers with clean energy vs. fossil fuels

To better understand the potential impacts of meeting US electricity demand growth from AI data centers, UCS’s new analysis highlights the consequences of data centers’ growing electricity demand on the US power grid and how decisionmakers can mitigate harmful economic, climate, and health impacts through clean energy deployment and strong ratepayer protections.

Using the National Renewable Energy Laboratory’s (which the Trump Administration recently renamed the National Laboratory of the Rockies) open-source  Regional Energy Deployment System’s (ReEDS) power sector model, we analyzed the impacts of meeting electricity demand growth from AI data centers on electricity generation, electricity costs, and the health and climate costs of emissions from using fossil fuels—as well as the avoided health and climate costs of emissions from using clean energy instead. Since future data center demand growth and the policy landscape are highly uncertain, we modeled multiple demand growth and energy policy scenarios.

Data center demand is surging—and highly uncertain

We project overall US electricity demand to increase 62% between 2025 and 2050 under a “mid-level” data center demand growth scenario, and as much as 79% under a “high” data center demand growth scenario (Figure 1). Data centers account for about half of total demand growth over the next five years, but this share falls over time as electrification increases in other sectors (especially transportation). While we vary the amount of data center demand growth, we assumed the same level of demand growth in other sectors under all scenarios. Studies by UCS and others show that higher levels of electrification of other sectors would likely be needed to achieve economy-wide net zero emissions by 2050.  

The number of data centers anticipated to be built in the United States and how much
electricity they will need are both highly uncertain. Recent actions by electric
utilities and regional transmission organizations indicate that many proposals to build data centers are redundant. In addition, utilities, which earn a guaranteed return on investment, have an incentive to overestimate future demand. This could lead to overbuilding the electricity system, higher costs, and stranded assets. A lack of transparency compounds the uncertainties, as proposals to build data centers and the power plants to serve them, are typically confidential.

Current policies increase reliance on fossil fuels

Under current state and federal climate and energy policies, our analysis finds that the United States is likely to increase its reliance on fossil fuels as electricity demand grows due to data centers and electrification of other sectors. We found that gas generation could increase 23% between 2026 and 2035 and 77% by 2050, with 90 gigawatts (GW) of new gas capacity added by 2035 and a whopping 335 GW by 2050. (One gigawatt is enough to power about 750,000 average US homes per year and could be supplied by 294 utility-scale land based wind turbines, according to DOE)

While our analysis projects coal generation to continue to decline in response to planned retirements, remaining coal plants are dispatched more to help meet the growth in demand. While the model includes the Trump administration’s rollbacks of EPA power plant carbon standards, we didn’t account for other regulatory rollbacks and subsidies that could further boost coal generation. For example, we didn’t model recent Trump Administration actions to use emergency powers to prevent at least six uneconomic coal plants from retiring, which are expected to increase electricity costs for consumers.

Even with these caveats, we found that the additional fossil fuels burned to power data centers could increase power plant carbon dioxide (CO₂) emissions by 19% and result in $1.6 trillion in climate and health damages over the next decade, growing to $4.5 trillion by 2050, under our mid-level data center demand growth scenario. We expect that recent actions not included in the model would only increase these costs compared to our results.

Stronger policies ensure that clean energy powers data centers

Restoring federal clean energy tax credits accelerates the deployment of wind and solar to power data centers. It also reduces reliance on gas and coal generation to meet the growth in electricity demand.

Adopting policies to reduce US power sector emissions of CO₂ would go even further, facilitating the clean energy transition to help meet economywide emission reduction targets. With a target of a 95% reduction in CO₂ below 2026 levels by 2050, wind and solar generation nearly triple between 2026 and 2035 and increase five-fold by 2050 to replace coal and gas. Combined, all renewable energy sources provide more than 60% of total US electricity generation by 2035 and 81% by 2050.

Data centers put ratepayers at risk of higher electricity costs

Rapid data center growth results in more than $500 billion in cumulative electricity costs nationally by 2035 and nearly $1 trillion by 2050 across our three policy scenarios and assuming mid-level demand growth. This represents 18-24 percent of total US wholesale electricity costs over the same period (see Figure 5). Without strong ratepayer protections requiring Big Tech companies to pay their fair share, these costs could get passed onto households and other businesses.

Restoring federal clean energy tax credits to the timelines and phaseouts that were included in the Inflation Reduction Act results in cumulative savings of $248 billion (4 percent) through 2050 compared with the “Current Policies” scenario. This underscores that recent actions by Congress and President Trump to rollback tax credits for solar and wind as part of the One Big Beautiful Bill Act (OBBBA) will make electricity less affordable, which is consistent with other recent studies.

Restoring the tax credits and adopting more ambitious climate and clean energy policies that nearly decarbonize the power sector by 2050 (the Low-Carbon Policy scenario) results in higher costs of $291 billion (12 percent) from 2026 to 2035 and $412 billion (7 percent) through 2050 compared with Current Policies. But these costs of transitioning to clean, low carbon energy sources are far outweighed by the avoided climate and health damages from burning fossil fuels, reaching more than $1.6 trillion between 2026 and 2035 and $13.3 trillion by 2050. Additional interventions, such as using clean electricity to replace fossil fuel use in other sectors, investing in energy efficiency, and increasing data center flexibility, could significantly lower overall energy bills.

We can avoid the health and environmental harms associated with unmitigated growth of data centers

Ensuring that data centers are efficient, flexible, and powered with clean energy requires stronger state and federal policies. Policymakers and regulators should require utilities and data center developers to be more transparent and accountable, improve long-term planning for meeting data center demand, and protect other customers from cost increases and negative health impacts.

They should also require utilities and data center developers to meet demand growth with new low carbon or zero carbon generation and use energy storage instead of diesel or gas generation for back-up power. Policies supporting the development of data centers should include standards and guardrails that protect public health while reducing emissions, energy and water use, and other environmental impacts.

Only bold action will ensure that the nation meets electricity demand growth with clean energy, achieves its climate goals, and protects consumers from added costs brought on by the growth of data centers. Unmitigated data center growth should not continue in the absence of stronger policies and guardrails, given that the consequences for energy affordability, reliability, public health and the climate are so high.  The path for the United States to achieve these benefits is clear.

Las consecuencias pueden ser peligrosas, como por ejemplo cuando las personas se ven obligadas a mantener la temperatura de sus hogares a niveles inseguros durante las olas de calor o las tormentas invernales debido al alto costo de la energía.

¿Cuáles son las causas de la crisis de la asequibilidad energética en Massachusetts?

Según un estudio reciente del Laboratorio Nacional Lawrence Berkeley, el precio promedio de la electricidad residencial aumentó un 27 por ciento entre 2019 y 2024 a nivel nacional.

En Massachusetts, el precio de la electricidad aumentó un 50 por ciento durante la última década. Son muchos los factores que han causado este aumento, entre ellos la inflación, pero una red eléctrica obsoleta y la dependencia excesiva del gas metano para generar electricidad en el estado han desempeñado un papel clave en los precios elevados que deben pagar los usuarios residenciales.

Las mejoras en los sistemas de transmisión y distribución de la electricidad por parte de Eversource y National Grid contribuyeron a un aumento de más del 50 por ciento en los cargos de distribución y de más del 70 por ciento en los cargos de transmisión durante los últimos seis años, según un análisis del Boston Globe.

Además, debido a que más del 50 por ciento de la electricidad de Massachusetts y de la región proviene de la quema de gas, se ha maximizado la exposición del estado a la volatilidad de los precios del gas.

Con el fin de enfrentar esta crisis de asequibilidad energética, los legisladores del estado están considerando diferentes mecanismos para reducir los costos de la energía. Lamentablemente, el último proyecto de ley de la Comisión de Energía de la Cámara de Representantes incluye disposiciones perjudiciales que retrasarían la acción climática y los avances en la adopción de la energía limpia.

El calor extremo y los picos en la demanda de electricidad durante el invierno señalan porqué es imprescindible centrar la acción climática en toda discusión sobre cambios al sistema energético del estado, incluyendo sobre la asequibilidad energética. Además, las energías limpias, como la energía solar, la energía eólica y la eficiencia energética, suelen ser las opciones de menor costo. Massachusetts tiene un enorme potencial para generar electricidad usando la energía solar y la energía eólica marina sin incurrir en costos adicionales por combustible. Las políticas estatales lo deben de reflejar.   

El proyecto de ley actualmente está siendo analizado por la Comisión de Medios y Arbitrios de la Cámara de Representantes. Si bien existe una serie de políticas y programas que podrían ayudar a reducir los costos de la energía, hay tres áreas que la legislatura debe priorizar para proteger eficazmente el bolsillo, la salud y el bienestar de sus electores, tanto ahora como a futuro.

1. Invertir en energía solar distribuida y eficiencia energética

Los recursos energéticos distribuidos (DERs, por sus siglas en inglés), tales como la energía solar en tejados, reducen el costo de la energía al disminuir la “demanda pico”, minimizar la necesidad de realizar costosas inversiones en sistemas de transmisión y distribución y reducir los precios de la electricidad en el mercado mayorista.

De hecho, durante un evento pico en el que se alcanzaron temperaturas de 100°F (38°C) en junio del 2025, un estudio del Acadia Center halló que los sistemas solares distribuidos (ubicados en el lado del consumidor de los medidores de electricidad de las empresas de servicios públicos), representaron un ahorro para los consumidores de al menos $8,2 millones en uno de los días más costosos del año para la red eléctrica. Los DERs también permiten que nuestro sistema eléctrico sea más confiable y sustituyen la generación de energía a partir de combustibles fósiles contaminantes.

La eficiencia energética y la electrificación también son fundamentales para reducir los costos que pagan los hogares y las empresas, así como disminuir la contaminación por combustibles fósiles. Por ejemplo, el programa de eficiencia energética del estado (MassSave) ha ayudado a los residentes a ahorrar más de $3.400 millones en sus facturas de energía entre 2012 y 2023.

El proyecto de ley sobre la asequibilidad energética debe reforzar el apoyo estatal a la energía solar en tejados, el almacenamiento de energía y la eficiencia y la electrificación, especialmente en las comunidades marginadas donde personas de bajos recursos y de comunidades étnicas y raciales diversas viven, y para quienes la asequibilidad y la resiliencia energética pueden tener aún un mayor impacto.

2. Proteger los compromisos del estado para enfrentar el cambio climático

Nunca antes había sido tan importante adoptar medidas continuas, explícitas y ejecutables para reducir nuestra dependencia de los combustibles fósiles. Massachusetts ha establecido unas metas climáticas bien pensadas y basadas en la ciencia, y debería ser líder en su cumplimiento, especialmente ahora que el gobierno federal ha renunciado a sus responsabilidades. Cualquier intento por dar marcha atrás a esos compromisos sería un paso vergonzoso en la dirección equivocada.

No podemos permitir las narrativas falsas que intentan enterrar la evidencia científica de los impactos y los costos económicos del cambio climático y que nos condenan a pagar costos más elevados por nuestra dependencia continua de los combustibles fósiles. El estado ya depende excesivamente del gas metano, un combustible con alta volatilidad de precios. La quema de combustibles fósiles para generar electricidad también intensifica los días de calor extremo, lo que provoca un mayor consumo y, por consiguiente, un encarecimiento de la electricidad.

Además, el impacto de esta dependencia va más allá del costo de la energía. Los desastres climáticos extremos en Massachusetts agravados por el cambio climático ya han tenido un costo de cientos de millones de dólares.

Es esencial ser audaces en nuestras acciones climáticas para avanzar la economía de energía limpia del estado, mejorar la salud pública y protegernos del agravamiento de los impactos climáticos. De hecho, una investigación realizada por Potential Energy Coalition entre más de 15.000 estadounidenses reveló que la gente está decididamente a favor de tomar acciones climáticas (66 por ciento) y promover la transición a la energía limpia (71 por ciento).

3. Evitar que las familias de Massachusetts subsidien a los centros de datos

Los centros de datos que se planean construir en el estado podrían requerir nueva infraestructura de redes eléctricas, y miles de megavatios de capacidad de generación eléctrica adicional, lo que representaría un aumento de casi un 25 por ciento. Sin embargo, el estado no cuenta con políticas para proteger a los contribuyentes de Massachusetts de los costos relacionados con dichos proyectos.

Un estudio reciente realizado por la Unión de Científicos Conscientes (UCS, por sus siglas en inglés) reveló que los contribuyentes de siete estados que son clientes de la organización de transmisión regional PJM tuvieron que pagar más de $4.300 millones por concepto de proyectos de transmisión relacionados con centros de datos, y que aún quedan miles de millones más por pagar.

Cualquier proyecto de ley sobre la asequibilidad energética debe requerir que los centros de datos paguen el costo de su demanda de electricidad, además de proteger a las comunidades de las consecuencias que implica la instalación de centrales contaminantes de gas y diésel que ponen en riesgo la salud. También debe requerir que la generación de electricidad para los centros de datos sea adicional y libre de carbono.

Legisladores de Massachusetts, contamos con ustedes para lograr una verdadera asequibilidad en el 2026

Estimados legisladores, por favor dejen a un lado las distracciones y eviten las falsas disyuntivas entre la asequibilidad por un lado y la energía limpia y la acción climática por el otro. Varias décadas de datos demuestran que los programas de eficiencia de MassSave, la implementación de la energía solar distribuida y otras tecnologías de energía renovable no son obstáculos para la asequibilidad energética, sino que son la base de la misma.

Las acciones continuas y ejecutables para avanzar las metas de descarbonización del estado nos protegen de los cada más graves y costosos efectos del cambio climático. Además, la protección de los consumidores frente a los centros de datos garantiza que los usuarios no paguen costos energéticos más elevados para satisfacer la demanda de los centros de datos y que se dé prioridad a la salud pública de los residentes al no permitir que la energía contaminante generada a partir de combustibles fósiles alimente a estos centros de datos. Contamos con ustedes para lograr una asequibilidad energética real ahora y en los años por venir.

Tu voz cuenta y puede hacer la diferencia

Pídele a tu legislador estatal de Massachusetts que tome estas medidas para proteger la asequibilidad energética y al clima.

Fair warning, this blogpost covers some pretty grim ground. But stick with me, please. Documenting the harms and injustices perpetuated by this administration now, as they occur, ensures we bear witness and that the hard work that made prior progress possible is not erased. Let’s make sure we don’t forget the important details as we fight to build a better, brighter future beyond this dark time.  A healthier, safer, and more equitable future is ours to create, as UCS President Gretchen Goldman says. 

Annus horribilis for people in the United States

From Day One, it was clear that this deeply anti-science administration was intent on blatantly furthering a fossil fuel agenda—people’s health and welfare be damned. President Trump has assembled around him an extremely unqualified, obsequious cabinet and set of advisors, most of whom have no dedication to the public interest and are instead devoted to doing his every bidding.

This increasingly authoritarian regime has operated with impunity to tear up climate and clean energy policies, lie about the scientific realities of climate change and the facts on renewable energy, and ram through measures to boost fossil fuels and the profits of polluters. They have attacked the federal scientific enterprise built up over decades through taxpayer investments, fired or forced out agency experts, and cut funding for critical science. And a compliant Congress has enabled this destructive agenda, including by rubberstamping some of the President’s illegal actions and by failing to exercise its constitutional powers to check his tyrannical power grabs. The passage of the OBBBA, with its multiple provisions directly aimed at undermining clean energy—including wind, solar, batteries, grid infrastructure, and energy efficiency—at the  President’s behest was a particularly egregious example of this.

The Sabin Center’s Climate Backtracker shows that, as of January 14, 2026, the Trump administration has taken nearly 300 actions to scale back or halt climate and clean energy progress. UCS’s six-month report on the Trump administration summarized many of the attacks on science and democracy as of July 2025. Many of these actions were previewed in the Project 2025 manifesto, but the magnitude of the harms, and the speed and intensity of the attacks, are shocking, and the impacts have been mounting.

Early destructive actions were taken by DOGE, spearheaded by Elon Musk, taking a hatchet to federal agencies tasked with protecting the public interest and advancing science and innovation. Subsequently,  Director of the Office of Management and Budget Russell Vought, an architect of Project 2025, has taken a personal and vicious role in many of these attacks (see here and here, for example).

And unfortunately, the full weight of the impacts on people and our economy are only going to become clearer this year, as words and cuts are translated into lived realities for communities across the country. At the same time, many of the administration’s unlawful actions are being challenged in court, and it has lost many of these cases, putting some brakes on some of its worst excesses.

Before diving further into details, it’s important to note two key themes: The Trump administration’s destructive actions are a direct threat to our health, our economic well-being, and to our nation’s ability to build a thriving, fair, innovative economy. These actions demonstrate an utterly corrupt government hell-bent on prioritizing the interests of polluters and billionaires over the needs of ordinary people.

While far from exhaustive, here are some of the major assaults on climate and clean energy from the Trump administration that we’ve seen in the last year:

1. Attacking agencies and organizations engaged in life-saving climate science research, data collection and monitoring

This has included threats to dismantle NOAA and NSF-NCAR; disbanding the author team for the sixth National Climate Assessment; taking down the US Global Change Research Program’s website, which includes all previous National Climate Assessments; and halting US federal scientists’ engagement with the Intergovernmental Panel on Climate Change (IPCC). NOAA, the nation’s foremost climate science agency, has faced reckless firing of staff, budget cuts, and slashed resources for climate research, satellite programs, data, and modeling.

Under Department of Commerce Secretary Howard Lutnick’s watch, the agency’s weather forecasting and climate monitoring capabilities are being undermined and many National Weather Service offices have been dangerously understaffed—undercutting critical resources that communities, first responders, farmers, mariners, businesses, and local decisionmakers rely on to protect lives, infrastructure, and economic activity. It’s crucial that the forthcoming Congressional appropriations process rejects the Trump administration’s budget proposals and restores healthy funding levels for federal science agencies.

2. Clawing back renewable energy funding and attacking clean energy projects

The administration has illegally frozen and clawed back billions in funding for climate and cutting-edge clean energy investments, including Department of Energy (DOE) grants and the Environmental Protection Agency’s Greenhouse Gas Reduction Fund. These actions have been challenged in court, and just this week a court has ruled that the Trump administration’s cancellation of some grants based on which states grantees are in violates the law.

DOE Secretary Chris Wright has repeatedly attacked clean energy, rolled back energy efficiency standards, overseen mass staff cuts at the agency, and renamed the world-renowned National Renewable Energy Laboratory to strike “renewable energy” from its name.

In addition to the major harms to clean energy inflicted by the OBBBA, the administration has also repeatedly and arbitrarily intervened in the leasing and permitting of a huge range of renewable energy projects, including pausing offshore wind, onshore wind, and solar projects. Wind and solar developers have just sued the Department of the Interior and the Army Corps of Engineers for “pursuing a concerted and illegal strategy to choke the ability of private developers’ ability to build new and much-needed energy generation projects.” Numerous attacks have been lobbed at offshore wind projects—including projects that were nearly completed—most recently by citing bogus “national security” considerations. The administration just suffered a major setback in a case brought by Revolution Wind, a project of Ørsted, a Danish offshore wind developer. These attacks on renewable energy have been accompanied by a raft of disinformation spouted by the President and his administration, contrary to the facts about the tremendous economic and health benefits of renewable energy, including solar and wind.

3. Gutting pollution standards and boosting fossil fuels

Environmental Protection Agency (EPA) administrator Lee Zeldin has launched an all-out assault on regulations, guidance, and scientific research aimed at protecting public health and the environment.  This has included weakening, rescinding, or delaying EPA regulations to limit heat-trapping pollution from power plants, vehicles and the oil and gas industry, as well as giving exemptions to polluters causing toxic air pollution from coal-fired power plants and other industrial sources.

The agency is also in the final stages of a process to overturn the science-based Endangerment Finding, a bedrock legal determination establishing the health-harming impacts of heat-trapping emissions. And in a major departure from precedent and long-standing best practice, EPA is also moving away from quantifying the public health impacts— including lives lost or saved—associated with agency rulemakings, debuting this egregious practice in a just-released rule for NOx pollution standards for new gas turbines. This alarming action is a complete capitulation to polluter interests and upends the agency’s mission to protect public health and the environment.

They’ve also taken unprecedented steps to cut the public out of the process of weighing in by eliminating the customary notice-and-comment period for regulations they arbitrarily designate ‘unlawful,’ and a sweeping executive order essentially allows agencies to wipe off whatever regulations they want to—as well as enforcement of those regulations in the time between.

The administration has also launched multiple direct attempts to boost coal, oil, and gas use under the guise of a spurious “national energy emergency.” DOE Secretary Chris Wright and Department of Interior Secretary Doug Burgum—both with deep fossil fuel industry ties—have aggressively embraced the president’s fossil fuel agenda. Burgum’s actions have included expanding oil and gas leases on public lands, rescinding requirements for environmental impact statements, and fast-tracking permits for fossil fuel energy, all while repeatedly interfering to stop deployment of renewables. The administration’s latest shocking move in this vein is its imperial and illegal grab for Venezuelan oil.

4. Attacking FEMA’s disaster response capabilities and investments in climate resilience

Department of Homeland Security Secretary Kristi Noem has overseen the firing and forcing out of more than 20 percent of FEMA’s staff already, with further steep cuts to FEMA’s “Cadre of On-Call Response/Recovery Employees” (CORE) expected imminently. At various points last year, Secretary Noem and President Trump went as far as calling for FEMA’s abolition! The agency has faced unending turmoil and dysfunction, with a series of unqualified acting chiefs quitting or being fired only to be replaced by yet another poor choice. This chaos led to major gaps in responding to disasters like the Texas flash flood last year. As the GAO points out, staffing shortages at FEMA have serious consequences for the agency’s ability to do its job to help communities hit by disasters.

The Trump administration also illegally cancelled FEMA’s Building Resilient Infrastructure and Communities (BRIC) program grants for states, which a coalition of twenty states took the administration to court over, and recently won their lawsuit. The Trump administration has taken actions that will leave communities less prepared and more at risk from worsening climate impacts, including rescinding the science-informed federal flood risk management standard, disbanding expert advisory councils, and politicizing and delaying disaster aid for states. My colleague Shana Udvardy has been carefully tracking the attacks on FEMA, including the recent last-minute cancellation of a recent FEMA Review Council meeting where they were supposed to release a report with recommendations, which has been under threat of interference from Secretary Noem.

5. Taking down or altering climate-related websites and datasets

This includes taking down the US Global Change Research Program (USGCRP) website and all the previous National Climate Assessments; taking down climate.gov, a free public portal for essential information on climate science and impacts (some of the information is now being curated at climate.us); removing climate science information from the EPA website; proposing to discontinue the GHG Reporting Program and datasets (like NOAA’s billion-dollar weather and climate-related disasters dataset, and its snow and ice data products); and failing to release the EPA’s Annual GHG Inventory for the United States (which the Environmental Defense Fund (EDF) successfully retrieved via a FOIA filing).

6. Lying about the facts on climate science

The most egregious example of this is the sham DOE “climate” report, which weaponized disinformation and uncertainty to downplay the risks of climate change and was invoked by EPA as part of its motivation for proposing to overturn the Endangerment Finding. This mirrors a classic strategy of employing disinformation and deception long practiced by the fossil fuel industry, now dangerously being adopted by the US government.

The Environmental Defense Fund and UCS have filed a lawsuit against the administration citing its violation of the Federal Advisory Committee Act (FACA) in the secret, illegal preparation of this report by five handpicked climate contrarians forming the Climate Working Group (CWG). The court has held that the CWG was an advisory committee subject to FACA and had ordered the administration to release all records related to its work. Other examples of this strategy include zeroing out the social cost of carbon (widely used as a measure of the monetary costs of climate damages caused by an additional ton of carbon emissions), directly ignoring the steep and mounting costs of climate change driven by fossil fuel emissions.

7. Withdrawing from international climate agreements and organizations

The most notable examples include withdrawing the UN Framework Convention on Climate Change, the Intergovernmental Panel on Climate Change (IPCC) and the Paris Agreement. The administration also single-handedly prevented the adoption of a major global agreement on reducing emissions from shipping, negotiated in the International Maritime Organization (IMO). Over 100 nations were on the verge of signing, but using bullying tactics—including direct threats to other nations—the administration succeeded in blocking and delaying this agreement.

Together, these actions underscore that this authoritarian, anti-science administration is determined to sacrifice people’s well-being and destabilize global cooperation. But forward-looking US states and the rest of the world recognize that devastating and costly climate impacts are mounting rapidly, and collective global action remains the only viable path to secure a livable future for our children and grandchildren. Withdrawal from global climate agreements and venues will only serve to further isolate the United States and diminish its standing in the world following a spate of deplorable actions that have already sent our nation’s credibility plummeting, jeopardized ties with some of our closest historical allies, and made the world far more unsafe.

As important as these individual attacks are, it’s crucial to also see the administration’s destructive strategy: They are trying to bury the evidence on climate change to advance a pro-fossil fuel agenda that delivers huge profits for a select few while the rest of us suffer the health and economic costs. And we can’t let them because the stakes are too high, for people today and for future generations. The stakes are especially dire for communities that have long been marginalized and discriminated against, those that bear the brunt of health-harming pollution and climate impacts and lack access to affordable clean energy and climate-resilient homes, here in the United States and around the world.

Looking back and fighting for a brighter future

Why look back on such a painful year? Because it’s a way to acknowledge and honor the people who were directly and harshly affected by the actions of this administration, including federal government scientists and frontline communities. Because remembering our shared history is how we build solidarity for the inevitable fights ahead. Because adversity teaches lessons. Because I believe the seeds of the destruction of this administration’s ill-conceived policies lie in their cruel overreach. Because we can take courage and inspiration from all the ways people across the nation showed up for our democracy, for science, for their communities.

This year has also brought extraordinary efforts to expose and fight back against the worst excesses of this unhinged administration. UCS has fought alongside many others by taking the Trump administration to court; advocating with members of Congress to stand up to the administration; filing technical comments with agencies; shining a light on the latest climate science and facts about clean energy; galvanizing the scientific community to get organized, join sign-on letters and call their elected representatives; uplifting the work of environmental justice experts; securing wins in states; joining nationwide demonstrations; and using our voice loudly in every venue we can to speak truth to power. 

And as we face down another tough year under the anti-science, authoritarian Trump administration, we’re fired up to keep up the fight for science and for our democracy. We hope you’ll join us—because despite it all, that future is ours to build.