In our continuing series following the Fiscal Year 2027 (FY27) appropriations process, we turn to the House Appropriations Committee’s Energy and Water Development (E&W) bill. This legislation contains the funding for two parts of the Department of Energy (DOE) of significance to the computing research community: the Office of Science (SC), home to most of the agency’s basic research support, and the Advanced Research Projects Agency-Energy (ARPA-E). After the Administration’s request called for steep reductions to parts of these accounts, the House has charted a markedly different course.

The House appropriators provide $8.52 billion at the topline for the Office of Science, a $120 million (1.4 percent) increase over the FY26 enacted level. Getting into the details, the bill funds the Advanced Scientific Computing Research (ASCR) program, home to most of the Office ‘s computing research efforts, at $1.18 billion. This is an increase of $60 million (5.4 percent) over the FY26 enacted level.

The committee’s legislative report provides specific policy direction within the Office of Science. It recommends not less than $245 million for quantum information sciences (QIS), divided between foundational research and the five National QIS Research Centers, with up to $15 million reserved for the Quantum User Expansion for Science and Technology (QUEST) program. It directs an Artificial Intelligence Workforce Development effort with universities, explicitly tied to the goals of the Genesis Mission, and orders a review of project management practices aimed at the cost growth and schedule slippage that have been squeezing SC’s research and operations accounts.

Within the ASCR portion of the report, the Committee concentrates its facility investment on the leadership computing enterprise, funding the new Equinox and Lux supercomputers at Argonne and Oak Ridge, sustaining the NERSC and ESnet user facilities, and supporting the High Performance Data Facility. It also provides not less than $310 million for applied mathematics and computer science research, including a small set-aside for aerospace-relevant quantum computing.

For ARPA-E, the Committee recommends $300 million, a $50 million (14 percent) reduction from the FY26 enacted level. As in last year’s bill, the appropriators offer little in the way of explanation for the reduction; the report’s ARPA-E section simply describes the agency’s mission and states the recommended figure.

Office of Artificial Intelligence & Quantum?
An interesting item in the House’s proposal is what is missing. Regular readers will remember that the Trump Administration proposed a new $1.2 billion Office of AI & Quantum (AIQ) in order to fund and operate the Genesis Mission. While Genesis itself runs throughout the bill, appearing to touch nearly every part of DOE’s mission, the House appropriators appear to be skeptical of the Administration’s and DOE’s proposal, and don’t provide funding for a standalone office.

Instead the Committee directs DOE to, “ensure that Genesis Mission initiatives align with the original purpose and intent of programs for which they are funded.” Additionally, the appropriators direct the department to develop and maintain a Department-wide strategy that aligns:

• AI and quantum research, development, test, and evaluation;
• Supports integrated discovery platforms combining advanced computing, large-scale datasets, AI models, and emerging quantum systems; and
• Promotes secure, energy-efficient AI deployment, including for high-performance computing and data centers.

It appears that the House appropriators want the coordination function of the AIQ office but without a new bureaucratic setup.

Indirect Costs
In the general provisions section of the E&W legislation, specifically section 312, the appropriators direct the Department of Energy to, “continue to apply the negotiated indirect cost rates…to the same extent and in the same manner…as applied in fiscal year 2024.” This language closely follows wording in the House’s CJS appropriations bill affecting NASA, NSF, and the Department of Commerce. The shift toward a more standardized approach across the different House Appropriations subcommittees, which handled the issue differently in last year’s funding bills, is a positive development that signals Congress is listening to the research community’s concerns.

Conclusion
The bill was approved by the full House Appropriations Committee on May 20. The legislation now heads to the full House chamber for approval. The next stage of the process is to wait for the Senate Appropriations Committee to release their version of the bill. However, it is unlikely that the FY27 process finishes by the beginning of the FY27 fiscal year (October 1). We are expecting a continuing resolution in the near-term, with any final resolution coming after the November Midterm elections and possibly into next calendar year. We will keep tracking the bill as it moves and post updates here on the CRA Policy Blog.

Core Elements of OMB’s Proposal

OMB is proposing to make substantial changes to what is known as the Uniform Guidance (specifically 2 C.F.R. Part 200). This is the single set of rules that governs the lifecycle of virtually every federal research grant, from how agencies review proposals, to how recipient institutions manage budgets and report results, to how auditors assess compliance. Any university department of computer science, any computing-focused research institute, and any industry lab holding a federal award is bound by it. Changes to the Uniform Guidance therefore have immediate, practical effects on how research is proposed, funded, conducted, and ultimately accounted for. Put another way, the Uniform Guidance is the framework that establishes the administrative requirements, cost principles, and audit standards that every university, research lab, and nonprofit must follow when managing federal awards from NSF, DOE, DARPA, NIH, and dozens of other agencies.

Key Changes that OMB Is Proposing

OMB frames the proposed revisions around three stated objectives: (1) improving transparency, accountability, and oversight; (2) clarifying the regulatory status of the 2 C.F.R. text; and (3) reducing recipient burden. In practice, the most significant changes fall into the following areas.

1. Elevating the Uniform Guidance from “Guidance” to Binding “Regulation”
   The current version of Part 200 explicitly states that it is “guidance, not regulation.” The proposed rule would erase that language and formally reclassify the document as an OMB regulation. The practical consequence is significant: once this change takes effect, future OMB amendments to Part 200 would apply government-wide on a single date, without requiring each funding agency to conduct a separate notice-and-comment rulemaking. Institutions would have less lead time and fewer opportunities for agency-specific input between the time OMB adopts a change and the time it binds them.
2. Political Appointee Review of All Discretionary Awards (§ 200.205)
   Perhaps the most structurally disruptive proposal is a new requirement that all discretionary awards, including competitive research grants, undergo pre-issuance review by a senior political appointee before they can be issued. The proposed rule is explicit that peer-review recommendations must remain “advisory” and may not be routinely deferred to or “ministerially ratified” by agency officials. The appointee must certify, among other things, that the award “demonstrably advances the President’s policy priorities” and does not fund activities inconsistent with administration policy.

   The rule also states that agencies are not obligated to issue an award simply because they published a Notice of Funding Opportunity (NOFO), they may re-post opportunities rather than fund proposals deemed insufficiently strong.

3. Indirect Cost Rates as a Competitive Factor (§ 200.205(b))
   Executive Order 14332 (August 2025) directed OMB to restrict indirect cost reimbursement for discretionary grants. Because Congress included language in the FY 2026 appropriations bills blocking a formal cap on negotiated facilities-and-administrative (F&A) rates, the proposed rule does not directly alter the existing indirect cost rate negotiation system. However, it would introduce indirect cost rates as a competitive factor in award selection: the pre-issuance review principles would require that, “all else being equal, preference for discretionary awards should be given to institutions with lower indirect cost rates.” This is particularly consequential for research-intensive universities, which typically have higher negotiated F&A rates that reflect the true cost of maintaining the infrastructure, such as laboratories, computing clusters, compliance offices, cybersecurity systems, that federal research requires.
4. New Nondiscrimination and DEI Restrictions (§§ 200.218, 200.219, 200.300)
   The proposed rule would codify into regulation several executive orders on DEI and gender ideology. Key provisions include:
   – Prohibition on disparate-impact liability (§ 200.218): Federal award funds may not be used to promote or support theories of disparate-impact liability, broadly defined as any framework that infers unlawful discrimination from statistical outcome disparities without evidence of discriminatory intent. An exception exists for internal research and program evaluation, provided federal award funds are not used for that work and results are not applied to award activities.
   – Prohibition on discriminatory event services (§ 200.219): Public entities receiving federal awards could not discriminate based on the “viewpoint, content, or subject matter of speech” in providing services for events held on their property, even when those events are not directly funded by the federal award.
   – DEI and gender ideology prohibition (§ 200.300): Federal agencies and pass-through entities would be required to ensure awards do not “fund, promote, encourage, subsidize, or facilitate” DEI or DEIA policies that violate federal anti-discrimination law, activities related to “gender ideology” as defined in the rule, or pediatric gender transition procedures for those under 19.
   OMB explicitly cautions that institutions should not assume prior practices will satisfy the new requirements.
5. Expanded Risk Assessment Factors (§ 200.206)
   Federal agencies would gain broad new latitude to evaluate applicant institutions during pre-award risk assessment. New permissible factors include an applicant’s history of “questionable practices” (plagiarism, non-replicable studies, conduct inconsistent with civil rights laws), affiliations with organizations that undermine public safety or national security, and, notably for universities, compliance with Section 117 of the Higher Education Act, which governs disclosure of foreign gifts and contracts. This last factor could intensify scrutiny of international collaborations and research partnerships at a time when computing research is inherently global.
6. Cost Allowability Changes — Publication, Conference, and Advertising Costs (§§ 200.432, 200.454, 200.461)
   The proposed rule introduces new prior-approval requirements for publication costs, conference costs, and fundraising costs, and would presumptively disallow advertising costs. These changes have direct implications for computing researchers, who routinely budget for conference travel, workshop organization, and open-access publication fees as core elements of their federally funded work.
7. Expanded Termination Authority and New Suspension Power (§ 200.340)
   Agencies would gain broader authority to terminate awards based on “the national interest” as assessed at the time of termination, rather than at the time the award was made. A new 90-day suspension authority would allow agencies to pause awards while weighing termination. A separate provision would permit federal agencies to cooperate with private parties pursuing claims against a recipient or subrecipient for failure to comply with award terms, potentially expanding exposure to False Claims Act liability.
8. Foreign Collaboration Restrictions (§ 200.220)
   The proposal expands restrictions on international collaboration and expands what has been called the “Wolf Amendment,” which has historically barred NASA from using funds to support activities that benefit China’s national space program, to cover additional situations. Given that computing research regularly involves international co-authorships, joint workshops, and collaborative grants, these provisions could be significantly disruptive.

What This Means for the Computing Research Community

CRA member institutions depend heavily on federal grants from NSF, DOE, DARPA, and related agencies to fund foundational work in AI, systems, security, human-computer interaction, and many other areas. Several aspects of the proposed rule present particular concerns:

• Politicization of merit review. Computing research has long benefited from a grant system in which proposals are evaluated on scientific merit by qualified peer reviewers. Requiring senior political appointees to independently review and approve discretionary awards — with peer review reduced to an advisory role — introduces uncertainty and potential delays into a process that institutions and researchers depend on for planning purposes. The proposed language around “anti-American values” and “Gold Standard Science” as review criteria is undefined and could be applied inconsistently.
• Indirect cost competitiveness. Research-intensive universities that are home to most computing research Ph.D. programs necessarily have higher F&A rates, reflecting investments in computational infrastructure, cybersecurity, research compliance, and graduate student training. Treating lower F&A rates as a preference factor in competitive review could disadvantage precisely the institutions that have the depth to conduct cutting-edge computing research.
• Chilling effect on international collaboration. Much of the computing research community is global. Expanded restrictions on foreign collaboration, combined with heightened scrutiny of Section 117 compliance during risk assessment, could make it harder to sustain productive international partnerships, attract international graduate students and postdocs, and participate in global scientific communities.
• Publication and conference costs. Computer science has an unusually strong culture of conference publication. Prior approval requirements for conference and publication costs would add administrative friction to research activities that are both scientifically important and standard practice in the field.
• Uncertainty around DEI-related research. Research on algorithmic fairness, bias in machine learning systems, broadening participation in computing, and the societal impacts of technology touches on areas the proposed rule targets. Depending on how agencies interpret and implement the new provisions, some computing research programs — including programs supported through CRA itself, such as efforts to broaden graduate student participation — could face greater compliance risk.

How to Submit a Substantive Public Comment

Public comments are a formal part of the rulemaking process and anyone has the right to submit comments as a private citizen (you can possibly do it in your professional capacity as well, though be sure to check with your employer first). When a final rule is issued, the agency must respond to substantive comments in the preamble. A well-organized record of thoughtful public comments can shape the final rule, provide the basis for legal challenges if the rule is adopted as proposed, and signal to policymakers the scope of concern in the research community. As of the publication of this article, over 37,000 comments have already been submitted.

The deadline to submit comments is July 13, 2026 at 11:59 PM eastern time. Comments submitted after this date will be considered only at OMB’s discretion.

Here is a step-by-step guide for writing and submitting an effective, substantive comment:

Step 1: Read the proposed rule or a reliable summary. The full proposed rule is 108 pages (91 FR 32198). At a minimum, read the executive summary and the section-by-section discussion of proposed changes. You can also read summaries from other credible groups, such as the Council on Governmental Relations’s (COGR) Policy Perspective, the Association of American Universities’s (AAU) resource page, a comparison dashboard by the Planetary Society, or this guest post explainer article on the Scholarly Kitchen. A plain-language summary is also available at regulations.gov.

Step 2: Identify the specific provisions that affect you. OMB has asked that each comment reference the relevant section number in brackets at the start of each substantive point — e.g., “[200.205]” for comments on the merit review provision. Focus your comment on the provisions you have direct experience with and can speak to concretely.

Step 3: Write your comment with specificity and evidence. Effective regulatory comments are not petitions or expressions of general concern. They are specific, evidence-based arguments about why a proposed provision is problematic or beneficial and, where possible, what a better alternative would look like. Consider including:

– A brief description of your role and how the rule affects your work (i.e. are you a faculty member, graduate student, department chair, research administrator, or industry researcher).
– Specific examples from your own experience (i.e. a project that would have been affected, a compliance burden that would result, a collaboration that would be jeopardized).
– Arguments tied to the regulatory text: what does the proposed language require, why is it unclear or unworkable, and what real-world consequences would follow?
– Proposed alternative language or approaches, if you have them.

Step 4: Submit through Regulations.gov. Go to www.regulations.gov/commenton/OMB-2026-0034-0001. Should you wish to submit comments, CRA recommends submitting them as a private citizen, though you can also submit as a group/organization (be sure to check with your employer first on any policies they may have). Comments become part of the public record, so do not include personal information you would not want published. You may also attach a file (PDF or Word document) in addition to or instead of typing your comment directly.

CRA’s Position

The CRA is closely monitoring this rulemaking process and collaborating with our partners throughout the research, science, technology, and STEM policy sectors. We intend to file formal comments opposing the proposed rule and are prepared to pursue further measures depending on the specifics of any final rule. We have also joined over 300 other organizations in writing to OMB to request an extension on the rule’s comment period; as of publication of this article, we have not received a response.

CRA believes that certain pillars are essential for the field to thrive: an unbiased merit review process, consistent indirect cost structures for research infrastructure, the freedom to form international scientific alliances, and the ability to study how computing affects society. While we support responsible federal funding oversight, any changes to the national research enterprise must be approached cautiously to minimize disruption. In our view, the implementation of this proposed rule would impose undue administrative burdens on researchers and organizations, ultimately undermining the very improvements OMB claims to seek.

Please continue to monitor the CRA Policy Blog for further developments and the most recent information regarding this topic.

First, let’s look at NSF. At the top line, the House’s FY27 proposal for NSF calls for a substantial cut, reducing the agency’s funding to a flat $7.00 billion. This is the same amount they proposed for the agency last year. This represents a cut of $1.75 billion, or a 20 percent reduction, from the FY26 level of $8.75 billion. While such a number is objectively bad, it is significantly better than the $3.96 billion requested by the President.

Looking at the Research and Related Activities (R&RA) account, home to NSF’s research portfolio, the account would receive $6.44 billion for FY27. That would represent a 10 percent decrease, or $740 million, compared to FY26 enacted levels.

However, there is a catch with R&RA’s number and it involves the Directorate for STEM Education (EDU). As they did last year, the House appropriators agreed with the Trump Administration’s proposal to fold EDU into the R&RA budget line. Since Congress has historically not stipulated funding at the directorate level, and continues to do so here, we do not have a specific budget number for EDU under the House’s plan. But to give an idea of the situation, if we combine the FY26 numbers for R&RA and EDU, it received $8.12 billion last year. Comparing that to the House appropriators’ number for R&RA means they are proposing a $1.68 billion cut, or 21 percent, to the combined accounts for FY27. That would likely mean that EDU would receive a significant cut in line with the Trump Administration’s FY27 proposal.

The Committee’s report, which outlines funding policy direction, continues the trend of sticking closely to the body’s proposal from last year. Once again, the House appropriators lead with “Maintaining American Leadership in Research” and recognize NSF’s critical role in scientific leadership. The Committee directs the agency to “prioritize research that aligns with vital national security priorities, including initiatives to advance AI and quantum computing.” Furthermore, it encourages the Technology, Innovation, and Partnerships (TIP) Directorate to be used to, “expand partnerships with the private sector through cooperative agreements and consortia that strengthen the domestic science and technology ecosystem.” Additionally, the Committee praises NSF’s “significant investments” in high-performance computing and artificial intelligence, urging continued focus on workforce development. Finally, the Committee also recommends at least $30 million for NAIRR in order to move the program beyond the pilot stage.

One matter that the appropriators do not mention is the fate of the Directorate for Social, Behavioral, and Economic Sciences or SBE. Regular readers of the Policy Blog will remember that the Trump Administration proposed eliminating most of SBE in their FY27 budget request. As mentioned with regard to EDU, it is normal for Congress to not direct funding at the R&RA level. However, it is troubling that the House is silent on the matter entirely. Without clear Congressional direction and intent, this potentially gives the Trump Administration a great deal of discretion on the matter. We will have to wait and see if the Senate includes language covering the SBE issue in their proposal. As a reminder, CRA led a joint letter, signed by a coalition of 38 STEM organizations, to both Congressional Appropriations Committee urging them to preserve SBE and the research it supports.

Turning to the National Institute of Standards & Technology’s (NIST), the standards agency fairs badly under the House’s plan. The committee proposes a 30 percent decrease, or a loss of $550 million, for the research agency’s topline in FY27. The agency’s budget would go from $1.85 billion in FY26 to $1.30 billion in FY27.

The Science and Technical Research and Services (STRS) account, which contains the bulk of the agency’s research portfolio, is slated for a 20 percent reduction, dropping from $1.25 billion in FY26 to $1.0 billion in FY27. This figure includes $275 million designated by House appropriators for “Scientific and Technical Research Projects,” which are one-year Congressionally directed projects, more commonly known as earmarks. When these earmarks are removed from the calculation for both fiscal years (the FY26 budget contained $273 million for such projects), in order to provide a more direct comparison, the actual year-to-year cut for STRS research funding would be 26 percent.

In terms of policy direction for NIST, the House appropriators emphasize the agency’s important role with standards around AI and quantum information sciences (QIS). The committee provides policy direction on a range of topics covering “Open Source AI Safety,” “Agentic AI Standards,” and “Quantum Cryptography.” Additionally, the committee directs NIST to spend no less than last year’s amount on QIS efforts. Finally the committee provides $15 million for the Center for AI Standards and Innovation (CAISI) to, “advance artificial intelligence research, standards development, and testing capabilities,” and directs NIST to administer a grant program for, “hardware security and secure location verification for AI chips,” in order develop and publish voluntary standards and guidance.

Turning to NASA, the space agency’s total budget would remain flat at $24.4 billion in FY27, matching the FY26 level. Within this flat topline, the agency’s Science account faces a significant reduction. At $6.00 billion, the proposed funding is a substantial cut of $1.25 billion, representing over a 17 percent decrease from the FY26 enacted levels. This is the same amount the committee proposed last year for NASA Science.

Finally, in the general provisions section of the CJS legislation, specifically section 542, the House appropriators direct the Department of Commerce, NSF, and NASA to, “continue to apply the negotiated indirect cost rates…to the same extent and in the same manner…as applied in fiscal year 2024.” The Trump Administration has attempted to cap indirect cost rates, also known as facilities and administration (F&A) costs, at almost all of the federal research agencies since coming into office. Congress has generally blocked their attempts, as they are doing here. This is a positive development for the nation’s research community and an improvement over how the House Appropriations Committee handled the matter in last year’s CJS bill.

Conclusion

The full House Appropriations Committee approved the CJS bill on May 13, and the funding legislation now awaits a vote on the House floor. We are now waiting for the release of the Senate Appropriations Committee’s funding plan for these agencies, though the timing for that remains uncertain. Within the S&T policy community, there is a general consensus that the FY27 budget will not be finalized by the start of the fiscal year on October 1. In fact, we aren’t expecting any final action on the budget until after the November Midterm elections, with a resolution more likely sometime in the 2027 calendar year. Please continue to check back on the CRA Policy Blog for the latest updates.

First, let’s look at NIST but first some context. Over the past several funding cycles, the agency’s budget has been quite difficult to assess because Congress has used it for a large number of Congressionally directed funding (ie: earmarks). Given most of that funding is for only one year, it makes a year-to-year comparison more difficult. But to provide some context, the agency had $660 million, spread across the agency, in Congressionally directed funding in the final FY26 budget. This money is zero’ed out by the agency in its request, which is a typical practice. In the chart below, we are comparing year-to-year top line budget numbers.

The top line budget request for NIST is $854 million, a decrease of $996 million from FY26, or a cut of 54 percent. The institutes’ Science and Technical Research and Services (STRS) account, where the majority of the agency’s research is housed, would see a decrease of 42 percent; going from $1.25 billion in FY26 to $729 million in FY27.

Within the agency’s justification document, the administration says the proposed STRS budget, “protects investments in critical, emerging, and digital technologies — and are aligned with the Administration’s priorities within artificial intelligence, quantum information science and technology, biotechnology and biomanufacturing, next-generation communications, and cybersecurity.” However, a closer look at the details reveals something different: the Critical and Emerging Technology Measurement and Standards (CET) program is the only portion of STRS slated for a funding increase. And while CET includes subprograms for Artificial Intelligence, Quantum Science and Technology, and Biotechnology and Biomanufacturing, only the AI subprogram would see more funding under the President’s budget plan; the remaining two would face cuts compared to their FY26 levels.

The next agency we look at is NASA. Under the President’s plan, the space agency would receive a 23 percent cut, going from $24.4 billion in FY26 to $18.8 billion in FY27; that’s a reduction of $5.6 billion. NASA Science, which handles the research funding at the agency, would see a much more significant cut of 46 percent. It would go from $7.25 billion in FY26 to $3.89 billion in FY27, a reduction of $3.36 billion.

Looking into the details of the agency’s request, only the Exploration budget line would receive an increase for FY27, reflecting the administration’s prioritization of human space flight. All of the other budget lines in NASA’s request, including Science, receive cuts of varying degrees. In the Trump Administration’s main budget justification document, it says the NASA Science reduction, “terminates over 40 low-priority missions to transform,” the program, “into one that is more focused and fiscally responsible.” The document uses the Mars Sample Return mission as an example of a terminated program, calling it “grossly overbudget.” Additionally, the agency is recommending the elimination of the STEM Engagement budget line and all programming, saying, “NASA’s primary role is space exploration,” and the agency, “will inspire the next generation of explorers through exciting, ambitious space missions.”

Finally, we come to the National Institutes of Health. Under President Trump’s plan, the agency would go from $48.72 billion in FY26 to $41.16 billion in FY27, a decrease of $7.56 billion or 15.5 percent. Meanwhile, ARPA-H, or the Advanced Research Project Agency, Health, would receive a 47 percent reduction to its budget, going from $1.50 billion in FY26 to $945 million under the proposed FY27 budget plan.

In NIH’s executive summary justification, the agency paints a bleak picture of what the agency’s request might do to their research community. The number of competing awards is expected to fall by 47 percent, while the success rate is expected to drop to 7.8 percent (it had been 14.9 percent in FY26).

As for ARPA-H, it is important to note that while the agency sits within the organizational chart of NIH, it operates as an independent agency. So its budget justification is separate from the parent agency. It says its budget, “supports advanced R&D aligned with HHS priorities and the Make America Healthy Again…agenda.” That mission is further organized across five focus areas: Addressing Chronic Disease, America-Made Manufacturing & Rural Access, Proactive Approaches to Healthy Well-Being, Healthcare Security, Efficiency, and Transparency, and American Leadership in Frontier Health Technologies.

As with the other research accounts we’ve profiled, it’s worth remembering that these are not final budgets and both chambers of Congress will have their say in the process before final numbers are set into law. We are already expecting that Congress will not finish FY27 on time, which is September 30th. In fact, the expectation here in Washington is the budget won’t be finalized until after the November Midterm elections at the earliest, and likely not until the 2027 calendar year. And the outcome of that election will heavily influence how any FY27 budgets look.

Next steps in the FY27 budget process are for each chamber of Congress to come up with their individual funding plans. That process is already in full swing in the House and we expect to receive our first indications from the Senate sometime in the summer. We’ll have updates as those bills roll out; keep checking back for more information.

A brief note on terminology for observant readers: we are using the Department of Defense nomenclature rather than the administration preferred Department of War. This reflects the legal reality that the authority to name federal departments resides with Congress.

Before we get into the budget numbers, a little refresher about the DOD research accounts: the DOD’s Science and Technology (DOD S&T) program is made up of three types of funding: 6.1 (basic research), 6.2 (applied research), and 6.3 (advanced technology development). These classifications are made up of individual accounts covering each of the four services, Army, Navy, Air Force, and Space Force, as well as a Defense-Wide classification. The Defense Advanced Research Projects Agency (DARPA) is a part of the Defense-Wide account.

Two of the three DOD S&T’s accounts do badly under President Trump’s budget plan. Basic Research (6.1), which is the main Defense Department supporter of fundamental research at US universities, gets hit with a 9.9 percent cut. The funding will go from $2.47 billion in FY26 to $2.14 billion in FY27 under the administration’s plan, a reduction of $230 million. The details within the 6.1 accounts are quite alarming: the Navy University Research Initiative (URI) subaccount is marked for elimination and the Army and Air Force subaccounts are cut at 20 and 9.6 percent, respectively, compared to their FY26 levels. Only the Space Force’s URI is relatively spared with a 1 percent cut. While there is no URI subaccount within Defense-Wide, that portion of the DOD’s overall 6.1 research funding would remain flat under the administration’s proposal.

The Applied Research (6.2) account is hit even harder, receiving a 15.5 percent cut; going from $7.21 billion in FY26 to $6.09 billion under the Trump Administration’s FY27 framework, a loss of $1.12 billion.

In terms of good news, Advanced Technology Development (6.3) would receive a significant increase, going from $12.20 billion in FY26 to $16.94 billion in FY27, an increase of $4.74 billion, or +39 percent.

DARPA also does quite well under the administration’s budget plan. The agency would see an increase of 15 percent, going from $4.37 billion in FY26 to $5.04 billion in FY27, an increase of $670 million.

What’s going on here? There are two ways of looking at these proposed budgets. Unfortunately, neither is encouraging and are quite troubling. The first perspective is the way the Pentagon leadership plans their budgets, which is from budget request to budget request. This comes from the simple reality that no military can plan a workable national defense strategy around the vagaries of a legislative budget cycle. When we look at it this way, FY26 requested budget to FY27 requested budget, the view is not good for basic research:

6.1 – reduced by 4.9 percent
6.2 – increased by 5.5 percent
6.3 – increased by 70 percent
DARPA – increased by 10 percent.

Going by this comparison, it can be inferred that defense basic research is not a priority for the Defense Department leadership and is being de-emphasized. This is further reinforced when we look further back, as 6.1 has seen a year-over-year decline in their requested budgets for the past several years.

The other way to look at these numbers is one we talk about almost every year: budget gamesmanship. Namely that money is pulled by DOD leaders from what is seen as a Congressional priority (i.e. research funding) to put toward something else that does not have the same support. If the scheme works, Congress puts money back into R&D and the moved money “sticks” elsewhere in the DOD budget. It’s not a new strategy, as the last couple of presidential administrations have done it.

The obvious problem with this approach is what if Congress doesn’t put the money back? You get hard budgetary cuts that have happened in recent years. It only goes to show that to start at a bad budget request is a good way to end with a bad yearly budget.

The budget process now heads to both chambers of Congress for consideration. We should see the House’s proposal for the Defense Department soon, with the Senate’s plans released some time during the summer. The science policy community expects Congress to not finish the FY27 budget process on time (i.e. October 1). In fact, it’s expected that any final budget will not appear until after the November Midterm elections and more likely won’t be finalized until sometime in the 2027 calendar year. We’ll have to wait and see what happens before we know more. Please keep checking the CRA Policy Blog for more updates.

The signers of the letter represent the full breadth of the nation’s STEM research community. These organizations advised Congress that they consider the work of the SBE Directorate to be an indispensable and essential part of the nation’s overall research enterprise. Even though these organizations represent scientific and technological fields that primarily receive research funding from other parts of the foundation, they view SBE, and the research it supports, as a critical asset for advancing their own scientific missions.

The letter cautions that eliminating SBE would be self-defeating for the nation, as the “industries of the future” prioritized by the Trump Administration, such as advanced materials and manufacturing, biotechnology, AI, and quantum sciences, rely on understanding the human element to be impactful. The organizations point out that solving the nation’s most challenging modern problems requires the crucial data and insights that SBE-supported research provides.

Download the Letter

As with the request for the National Science Foundation, this year’s proposal is almost a carbon copy of last year’s terrible request for the Office of Science. And despite artificial intelligence (AI) and quantum information sciences (QIS) being administration priorities, even standing up a new office, the President is calling for significant cuts to research generally and a slight cut for computing research specifically.

The FY27 request for DOE SC is $7.14 billion, which is a decrease of $1.26 billion, or 15 percent, compared to the approved FY26 level of $8.40 billion. According to the SC’s budget justification documents, the requested budget goes to funding, “basic research to advance energy technologies, transform our understanding of nature, and strengthen the connection between advances in fundamental science and technology innovation.” Additionally, the office prioritizes Administration and DOE priorities, “emphasizing transformative advancements in artificial intelligence (AI), quantum information science (QIS), fusion energy, high-performance computing, and critical minerals and materials.”

Within the Office of Science account, the Advanced Scientific Computing Research (ASCR) program, home to most of SCs computing research efforts, would fare relatively well. The program would be funded at $1.10 billion, which is a slight cut of $20 million, or 1.8 percent, compared to FY26. As for the other SC programs, all received cuts of varying degrees, with Basic Energy Sciences (-20 percent) and Biological & Environmental Research (-54 percent) receiving the largest cuts.

Within the program’s budget justification section, it talks about, “ASCR’s leadership role in the design, delivery, and continuous improvement of the Genesis Mission platform, including the American Science Cloud,” as well as its role in, “expanding SC’s ability to leverage next-generation advanced computing, frontier AI models, and data to further research frontiers.” Further, the program’s request talks about ASCR’s basic research funding for applied mathematics and computer science, “to combine exascale computing and AI for innovation and next-generation computing paradigms,” in order to, “ensure U.S. leadership at the forefront of computing.” The language in the request also talks about ASCR’s role in R&D for QIS technologies, including quantum computing and networking, and next-gen distributed quantum computing systems, as well as next-generation user facilities in order to, “deliver integrated, high-uptime, extreme-scale HPC, data, and networking infrastructure for DOE science in the AI era.”

As for ARPA-E, the agency’s budget would be cut almost by half. Under the President’s budget proposal, ARPA-E would receive $200 million, a decrease of $150 million over last year, or a 43 percent cut. That is the same amount the Trump Administration proposed last year for FY26. The budget justification has few specifics, speaking only in generalities, saying the program will, “continue to fund and direct the discovery of outlier energy technologies that ensure the production of reliable, American-made energy.” It goes a little further, saying the agency’s request supports, “the Administration’s goal of restoring U.S. energy dominance through firm, baseload power,” and, “increase the energy available to power modern life and unleash American energy innovation to maintain America’s global competitiveness.”

Finally, DOE is proposing to stand up a new Office of Artificial Intelligence & Quantum (AIQ) to handle Genesis Mission activities. The office would be funded at $1.2 billion to, “implement the Administration’s objectives to advance bold, transformational leaps in U.S. science and technology (S&T) in computing for artificial intelligence and quantum research to ensure America remains the global S&T leader for generations to come.”

Office of Artificial Intelligence & Quantum

One of the outstanding questions that the science policy community had with this budget proposal was how much funding support the Administration would provide to the Genesis Mission. AIQ is the answer.

According to DOE, the new office will be the, “centralized coordinator and leader of all AI research, quantum activities, and Genesis Mission activities throughout,” DOE. However, from a computing research perspective, it’s unclear how beneficial AIQ might be. Within DOE’s budget documents, the department says that the office will, “support multiple AI supercomputers at Argonne and Oak Ridge National Laboratories and is also responsible for the Genesis Mission delivery, aligning investments, developing workforce, optimizing shared infrastructure and tracking performance.” Additionally, under “Program Highlights,” AIQ is tasked with supporting the facility operations and upgrades of new supercomputers at the national laboratories, and will, “explore incentive-based competitions to support the demonstration of scientifically relevant quantum computing.” All of which seems more aligned with infrastructure development, rather than the immediate support of research. Finally, the office’s proposed budget amount is suspiciously close to the amount that SC was cut by. While AIQ may prove beneficial to AI and QIS research in the long run, its immediate impact does not appear to benefit these fields.

Conclusion

As with last year’s budget request, while not as disastrous as the cuts to NSF and other federal research agencies, these funding proposals are still bad and undermine critical investments in U.S. research and innovation. CRA continues to oppose the President’s budget plan and encourages all members of the CS and IT research community to call the offices of their Congressional delegation to urge them to reject the Trump Administration’s budgetary framework for FY27. Our Advocacy Action Alert from last year still has relevant talking points that can be adapted for this year’s request.

What are the next steps? The process heads to Congress in order for both chambers to work out their own funding plans. We should start seeing the House’s proposals soon, with the Senate’s plans likely coming into focus some time during the summer. It is already expected that Congress will not finish the budget process on time (by October 1). In fact, any final FY27 budget is unlikely until after the November Midterm elections and more likely won’t be done until next calendar year. We’ll have to wait and see what happens before we know more. Please keep checking the CRA Policy Blog for more updates.

The Computing Research Association views with deep concern this decision by the president. This abrupt action is profoundly disruptive and threatens to undermine the stability and effectiveness of both the National Science Foundation and the nation’s scientific research enterprise. The NSB plays a critical, non-partisan role in guiding NSF and shaping national science and engineering policy. Its members are highly respected experts, chosen for their commitment to advancing the frontiers of knowledge and maintaining America’s global competitiveness. To summarily dismiss these dedicated leaders sends a troubling signal regarding the independence and continuity necessary for long-term scientific planning.

In response to the release of the president’s budget proposal, CRA released a statement expressing disappointment and alarm for the proposal, and calling on the Administration and Congress to reconsider this path.

As a general note, it appears that NSF’s budget request was prepared before Congress finished its work on the agency’s FY26 budget. With that in mind, in the agency’s budget documents, you will see they compared the President’s request to FY25 numbers (ie: the last year they had approved funding). Since CRA now has most of the numbers for FY26, we will compare the President’s request against that.

There are exceptions, most notably the Research & Related Activities (R&RA) directorates. Congress does not typically fund at that level, leaving it to NSF to decide how funding is distributed. Since the agency has not released finalized directorate budgets for FY26, we can only do a FY25 to FY27 request comparison for those parts of the agency.

At a high level, under the Trump Administration’s FY27 plan, NSF would see a 55 percent cut compared to their FY26 budget approved by Congress. At the topline, NSF would go from $8.75 billion in FY26 to $3.96 billion in FY27, a decrease of $4.79 billion. R&RA, the subaccount that contains the funding for research grants, would see a similar sized cut, going from $7.18 billion in FY26 to $3.41 billion in FY27, a decrease of $3.41 billion, or -52 percent.

The STEM Education Directorate (EDU), which contains the agency’s education programs, would also see a similar cut of 54 percent. It would go from $938 million in FY26 to $428 million under the President’s FY27 proposal, a reduction of $510 million. Additionally, the Administration is again proposing that EDU be folded into the R&RA account. This was proposed last year and Congress rejected it.

The Administration is also proposing the elimination of the Social, Behavioral, and Economic (SBE) Sciences Directorate budget line and much of its operations. SBE is the primary funder for social science research at NSF. For the computing research community, the research funded by SBE is critical to several areas of CS and IT, including cybersecurity and human-computer interaction, among many others. More analysis and potential impacts are below.

Finally, it is not surprising that the agency’s proposed budget would also see large decreases in the number of grants the agency funds and the researchers it supports. NSF estimates that its requested budget would fund an estimated 2,100 research grants in FY27 (5,800 were funded in FY25), allowing for a 7 percent funding rate (it was 19 percent in FY25). The agency estimates that its activities will support a total of 94,100 people in FY27, a drop of almost two-thirds from FY25 numbers (305,400). That number includes senior researchers, other professionals, postdocs, graduate students, and undergrads.

CISE Directorate:

The Computer and Information Science and Engineering Directorate’s (CISE) would receive a significant reduction to its budget in FY27 under the President’s proposed budget. CISE, located within R&RA and home for most of the computing research support at NSF, would see its budget drop from $933 million in FY25 to $346 million for FY27, a decrease of $586 million or 63 percent. (NSF has not yet released the spend plan for CISE’s FY26 budget.)

The CISE Directorate will, “support research on critical technologies to fuel economic growth, unleash American prosperity, and bolster national security.” Additionally, the directorate will support, “research across subfields in the AI and computing ecosystem, including novel algorithms, hardware and systems, programming and software engineering paradigms, cybersecurity, advanced wired and wireless networking, robotics, and quantum and bioinspired computing approaches,” and their investments will also include, “advanced cyberinfrastructure within computing and across science and engineering, and novel educational approaches…to effectively train students and workers for the jobs of the future.”

It should not be a surprise that artificial intelligence is the first bullet within CISE’s highlights. Specifically, the directorate will, “(a)dvance AI through support for foundational and use-inspired research, education, and infrastructure,” which align with the priority areas identified in relevant Administration RFIs and the Executive Order on Advancing Artificial Intelligence Education for American Youth. The document also says that, “key investments,” within the AI space, “will include the National AI Research Institutes and the National AI Research Resource,” also known as NAIRR.

There is not a specific budgetary number for NAIRR that NSF is advocating for in their budget documents. NSF does say that this request supports, “transitioning…NAIRR beyond the pilot stage to drive US AI innovation and discovery.”

While no AI funding is listed in the CISE section, the “NSF Wide Investments” section notes a reduction for AI overall at the agency. NSF’s budget documents show AI funding dropping from $965.18 million in FY25 to $655.23 million in FY27, representing a cut of approximately one-third.

Looking at other CS/IT areas in CISE’s budget, the directorate will, “support research and education that advance the foundations of quantum information science.” It will do that using, “novel approaches for building the quantum computing stack, “hybrid approaches that combine quantum and classical computing,” and “high performance post-quantum cryptographic algorithms and implementations that are secure against attacks from both classical and quantum computers.”

Similar to AI, there is no specific amount allocated for QIS within the CISE section. Instead, there is an amount listed in the NSF-Wide Investments section. The budget documents indicate a planned reduction in the agency’s funding for QIS, mirroring the trend for AI. NSF spent $365.46 million on QIS in FY25, with a planned expenditure of $231.15 million in FY27.

Other major investments in the CISE request are support for, “other areas critical to U.S. technological leadership, such as advanced wireless communications technologies, advanced materials and manufacturing including semiconductors and microelectronics, biotechnology, and cybersecurity.” No other details on these matters are given in CISE’s section. However, there is a reference to others research areas that CISE investments support, such as, “advanced cyberinfrastructure within computing and across science and engineering,” and, “novel educational approaches that stay ahead of fast-moving and disruptive technologies, like AI and quantum, to effectively train students and workers for the jobs of the future.”

TIP Directorate:

The Directorate for Technology, Innovation, and Partnerships, or TIP Directorate, will not escape cuts under President Trump’s proposal. Though, like CISE, the exact dimensions of the cut are unclear, as we do not have a FY26 budget for the directorate. TIP requested a budget of $350 million for FY27, a 43 percent decrease from its FY25 budget, which was $618 million.

TIP says they are, “fostering a new era of American innovation,” by accelerating R&D through, “strategic investments in key technologies,” like AI, QIS, and biotechnology. The directorate is, “using new approaches that enable new entrants to engage in the innovation enterprise.” Ultimately, TIP says its mission is to serve, “as the bridge between foundational investments by other parts of NSF and applied research by a range of stakeholders, derisking breakthrough technologies to the point of uptake by private industry and other federal agencies.”

With that in mind, TIP views its FY27 strategy through three broad strategies:

Strategy 1: Build a durable U.S. competitive advantage in critical and emerging technologies

This strategy aims to ensure that the United States maintains its leadership in sectors of significant economic and national security importance.

Strategy 2: Establish a new frontier of American innovation

NSF Engines awardees are dedicated to enhancing domestic capacity not only to generate ideas but also to rapidly bring them to the market. Ultimately, NSF is fostering a new wave of products and solutions conceived, manufactured, and consumed in America by American workers.

Strategy 3: Attract and train a world-class science and technology talent base and workforce ready for global competition

NSF is committed to creating accessible pathways into future jobs by investing in institutions and infrastructure responsible for training workers in current competitive technologies and emerging technologies that will define the global job market.

TIP’s budget section contains a list of selected investments and references their corresponding strategies. Unfortunately, even here NSF’s budget proposal includes cuts. For example, within the “Emerging Technologies Investments” under the first strategy, the directorate plans to reduce spending on artificial intelligence from $113.27 million in FY25 to $72.23 million in FY27. A similar reduction is proposed for QIS, with the planned spending dropping from $39.80 million in FY25 to $27.05 million in FY27.

SBE Directorate:

As previously mentioned, NSF is planning on eliminating, at least in part, the Social, Behavioral, and Economic Sciences Directorate. This action is not surprising, given the long-standing skepticism towards social science research, and SBE itself, from certain ideological groups in Washington.

However what is surprising is that news reports point to the agency pursuing this without receiving direction from Congress, or seeking input from the NSF community and stakeholders. This move is also being pursued despite the fact that in the FY26 funding legislation Congress specifically prohibited the Administration from reprogramming funding by elimination of a “program, project, or activity.”

There is one part of the directorate that is being preserved. Specifically, the National Center for Science and Engineering Statistics (NCSES) will, “operate independently of the directorates and continue to be supported through the R&RA appropriation.” As stated in NSF’s budget document, NCSES is one of, “the Office of Management and Budget’s 16 recognized statistical agencies and units in the Federal statistical system and has the statutory mandate to collect,” and report out, “policy-relevant information regarding the science and engineering enterprise in a global context.” However, even here NSF is proposing reduced funding, with NCSES’s FY27 proposed budget at $44 million, an over 40 percent reduction from FY25 levels ($74 million).

CRA is deeply concerned by the proposal to eliminate SBE and will monitor the situation throughout the budget cycle. SBE-supported research is essential to a number of computing and IT research subfields, including cybersecurity, human-computer interaction, and robotics. Furthermore, SBE-funded research will be indispensable for understanding the broader impacts of artificial intelligence on society.

Conclusions:

There are two important things for the computing research community to keep in mind with any presidential budget request. First, that this is only the first step in the process. Second, with the federal budget process, past success does not mean future success.

Congress will have their say in the process. In fact, the House Appropriations Committee has already announced an aggressive schedule of hearings and markups for all of the FY27 funding bills. And as shown with FY26 final funding levels, Congress does not share the Administration’s views on how to handle the nation’s research enterprise.

However, just because Congress rejected the President’s proposal last year, that does not mean they will do so again. In fact, we are expecting this year’s budget process to be especially long, given that the nation’s political attention is increasingly focused on the November Midterm elections. The expectation is that Congress will not pass final funding legislation until December, at the earliest, and more likely some time in the 2027 calendar year. The Fiscal Year 2027 budget starts on October 1, 2026.

With that in mind, the computing research community must keep Congress aware of the dire implications of the President’s proposed funding plan and that it should be rejected. We continue to encourage the entire CS and IT research communities to reach out to their Congressional delegations and explain how the Administration’s plans will impact their work and communities. You can find useful advice and talking points in our Advocacy Action Alert from last year; most of the talking points are still relevant with this budget proposal. As with last year, we hope that Congress will continue to take steps to protect the nation’s research enterprise from these proposals.

CRA will continue to engage with our Congressional friends and allies, and make the case with more skeptical members of the legislature that the nation’s research enterprise is worth protecting and advancing. We are closely monitoring the budget situation and will report any new developments to the community.

CRA Statement on the President’s Fiscal Year 2027 Budget Request

The Computing Research Association (CRA) expresses its profound disappointment and alarm regarding the President’s Fiscal Year 2027 budget request. For the second consecutive year, the Administration has proposed a budget that would inflict deep, damaging cuts on federal science agencies, threatening the vitality of the American research enterprise and our long-term global competitiveness.

While we recognize and share the Administration’s commitment to advancing “industries of the future,” such as Artificial Intelligence (AI) and Quantum Information Science (QIS), and applaud new efforts to prioritize those topics at the National Science Foundation (NSF) and Department of Energy (DOE), we are deeply concerned by the proposed budget cuts elsewhere for science in the President’s request, especially the deep cuts to NSF’s Computer and Information Science and Engineering (CISE) Directorate and the elimination of the Social, Behavioral and Economic Sciences (SBE) Directorate. To prioritize AI and Quantum while defunding CISE and dismantling the SBE Directorate is a misguided strategy that fundamentally misunderstands the nature of scientific and technological progress.

The CISE Directorate is the key supporter of fundamental computing research at the agency that bears the largest share of support for fundamental computer science research in the United States. Cutting funding for CISE while attempting to prioritize research in AI and QIS is like watering a single branch of a tree and neglecting the roots.

Like CISE, the SBE Directorate is not a luxury; it is a critical component of the nation’s scientific enterprise. Zeroing out SBE funding undermines the Administration’s own priorities, as pressing computing challenges require more than “pure” technical research. SBE-supported research is essential for understanding how humans interact with technology, and will be key for understanding such things as societal data use in AI algorithms, cognitive load, human trust in automation, economic impacts of labor displacement, and addressing the human behavioral roots of most cybersecurity breaches.

Undermining NSF and the interdisciplinary nature of modern science carries a steep cost. It cedes our leadership in the very technologies the Administration claims to prioritize and diminishes our ability to address complex national challenges.

CRA urges the Administration and Congress to reconsider this path. We must move toward a budget that reflects the reality of modern innovation: a holistic approach that integrates the study of human behavior and social systems with the development of the next generation of computing. Sustained, across-the-board investment in NSF, DOE, and the other federal science agencies is essential to ensuring America’s continued prosperity and security.