The bombshell part frontloaded in the Health Act 2026 (full 200 pages here) is the abolition of NHS England. The intent is to “put power and resources in the hands of frontline NHS organisations by abolishing NHS England and stripping back national bureaucracy”. The organization’s responsibilities would be devolved by the Secretary of State to any one or more of the following, largely at more local levels: 

• the Secretary of State
• an integrated care board (ICB)
• a company formed under section 223 of the National Health Service Act 2006
• a Special Health Authority
• an NHS trust
• an NHS foundation trust
• a Local Health Board
• any other public body

Local ICBs would be more numerous and gain more powers, but have more direction from and control by central government, via the DHSC.

The second bombshell is the Single Patient Record (SPR). The SPR would require all GPs, including private providers, and hospitals to share data on patients so that information that is now fragmented can be seen anywhere in England. This would redirect the information to the right doctors, nurses and specialists to securely see a patient’s full medical history. The Government claims this can go through as early as 2024 for maternity and frailty care.

The SPR is the means to create savings at the A&E and GP level. DHSC estimates that the combination of SPRs with virtual care would reduce A&E visits for frail patients by about 10,000 a year plus reduce another 10,000 visits due to fewer misdiagnoses. Other estimated annual savings would be for doctors’ hours–500,000–and 6,000 fewer hospital admissions, totaling £20 million. Patients would also have greater access to their health data including who may access it. It would join up community services with an audit trail to track anyone who accesses the record.

What’s disputed is controlling the data and supervising its security. At present, GPs are the data controllers for their patients’ records. They are permitted to share them with third parties for research purposes.  This would change to the DHSC. The British Medical Association (BMA) opposes this; their GP committee has warned that any move to take control of data away from GPs would damage trust and risk confidentiality.

Other parts of the bill reinforce NHS’ virtual hospital model through the NHS app. NHS Online is scheduled to launch in 2027 for planned specialist care. It’s estimated to provide the equivalent of up to 8.5 million appointments and assessments in its first three years. The other is that there is a “Duty to Promote Innovation” (section 6) by the Secretary of State, including payments and prizes. 

Effects on patient health and safety include the abolition of the independent patient advocate, Healthwatch, with its duties transferred to the ICBs and local councils. The Health Services Safety Investigations Body (HSSIB) will merge into the Care Quality Commission (CQC) regulator.

Effects on other sectors from the PinsentMasons analysis:

• On private healthcare and outsourced providers, more power will go to the ICBs and the DHSC.  The DHSC will have increased governance control through “the power to be able to cap day-to-day spending limits on NHS Trusts, appoint foundation trust boards, issue directions to ICBs regardless of performance, and may shift care provision between public and private sectors depending on whether or not doing so is in the interest of the NHS.”
• For life sciences and medtech, the focus on innovation is a major plus. The minus is that there will be more layers and structure to pass through.
• For health tech and data companies, the interoperability demands and innovation requirements are solid pluses. There will be “clear opportunities in platform, cloud and infrastructure provision, along with integrating AI and analytics systems for operations in clinical and administrative sectors.” Concerns are data governance and management.
• For social care and adjacent services, this reinforces the trend towards tighter integration. This will include “closer alignment with NHS commissioning structures, opportunities to participate more formally in integrated care pathways, and (sic) having to navigate increased and more centralised regulatory oversight.”

Implementation would be expected to start this year and extend over the next decade. Additional information from the NHS: press release, collection page for the Health Bill, Impact Assessment Summary. The Guardian

The Intellectual Property Office (IPO) launches the Knowledge Asset Management Hub (KAM). The KAM Hub’s purpose is to assist universities and other research organizations in identifying and commercializing their innovations, IP, and other knowledge assets. It was announced at the Knowledge Exchange UK Conference 2026 in Newport, Wales, by IPO Chief Executive Adam Williams. The Hub’s assistance spans four component areas: institutional IP strategy guidance, project-level IP risk and opportunity tools, patent data analysis and IP due diligence resources, and the Knowledge Asset Management Toolkit. UK.gov:  IPO release, KAM Hub document list

UnitedHealthcare is the largest provider of SCO plans in Massachusetts. SCO is a state Medicaid program that covers members 65 or older who live in designated areas of the state. Based on a mandatory in-home clinical assessment to determine their health status, covered members are assigned a level of care based on health conditions, with Level 1 being the least serious, Level 2 having behavioral health or substance use disorders, and Level 3 having the most serious health conditions. The levels have correspondingly escalating levels of payment to the plan.

The Commonwealth is stating that members were classified by UnitedHealthcare to Level 2 despite no other diagnosis or treatment. Other patients were assigned to Level 3 through improper assessments. Moreover, United allegedly knew that beginning in 2018 and continuing into 2019, members were improperly assigned to Level 3. United never disclosed the error nor repaid the higher payments resulting from the assignment. 

From the Office of the Attorney General release, “The AGO alleges that these were intentional failures, the result of a ‘growth at all costs’ strategy employed by United that incentivized and encouraged its field nurses to code MassHealth members as sicker or less able than they were.” UnitedHealthcare responded that the fraud allegations and the lawsuit are “meritless” and that “[The complaint] doesn’t accurately describe our Senior Care Options program, which helps seniors with complex care meet their individual health needs.”  MedCityNews, Healthcare Dive  

A nice ladder for Teladoc is the addition of their telehealth services to Walmart’s digital health offerings. Walmart’s Better Care Services will add Teladoc’s telehealth services in virtual urgent care for common conditions, dermatology, and nutrition for their customers. Payment can be either cash pay at $89/visit or through insurance. Prescriptions, if needed, can be sent to a pharmacy, including Walmart’s, which has same-day delivery  in as fast as an hour available in many locations. Walmart added Teladoc’s BetterHelp mental health services in January. Teladoc release, MedCityNews, Healthcare Dive

And funding ladders continue, indicating some freeing up of investment cash for companies that have gone through a few (or more than a few) rounds.

Glucose monitoring system Signos raised a $20 million round. The investment is from GV (Google Ventures), Dexcom, and Blue Cross Blue Shield of Alabama. Funding to date including this round is $57 million.  Signos is also a partner with Dexcom through their direct-to-consumer continuous glucose monitoring website, Stelo.com. Signos’ app for tracking glucose for weight management is FDA-cleared and includes insights into how behavior, lifestyle, sleep, and stress affect weight. It can also detect low and high glucose levels. Membership is $127/month including two sensors a month on the six-month plan. Mobihealthnews, Yahoo Finance

Healthcare provider data and directory company H1 raised $40 million, led by CVS Health Ventures. Across ten lettered and unlettered rounds, funding to date is $233.9 million. It has developed an AI-powered platform for identifying and engaging the right doctor across life sciences, payers, providers, and patients.  The H1 Doctor Graph platform is a structured representation of physician identity, expertise, relationships, and behavioral signals that identifies and engages the right doctors for critical workflows across pharma, health plan, health system, and technology companies. CVS Health and H1 have collaborated on several projects, including an AI model that improved the accuracy of their health care provider directory. Release, Mobihealthnews

It’s raining mega-IPOs.

The S-1 filing comes on top of their 28 May announcement of a $65 billion Series H funding led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. The valuation of $965 billion makes Anthropic the most valuable AI company on Planet Earth and perhaps the entire Solar System, surpassing OpenAI by about $113 billion. Anthropic’s valuation in February was $380 billion with their Series G raise, so the Series H valuation multiplied that by a stunning 2.5 times+ in three months. The new funds will be used for AI research, expanding its computing power for Claude, and scaling its products and partnerships.  CNBC, Mobihealthnews, Anthropic release

Anthropic’s over-the-top valuation was boosted by its projected annual revenue run of $50 billion, tipping into profitability this quarter, beating its own growth metrics regularly, and introducing new Claude products such as Claude Opus 4.8 and Claude Mythos Preview with advanced cybersecurity available to a limited group of companies. Anthropic in January rolled out Claude for Healthcare for providers and consumers. Claude pulled in front in corporate sales, ahead of OpenAI, as of April, and last month inked a new partnership with Bristol Myers Squibb to implement Claude throughout the company.

But are we at Peak AI? Axios, never one to shy away from Cold Buckets of Water when it makes for a good lede, has been trumpeting for the past week that companies are suddenly shying away from their “discovery” of soaring AI costs. Suddenly, ballooning IT costs, uncertain productivity gains, and a strange combination of employee overuse and sudden skepticism are causes for concern. An AI consultant told Axios that employees blew through half a billion dollars in a single month because they didn’t put usage limits on Claude licenses. Then the CEO of an AI software company, CloudBees, admitted that companies are using workforce cuts to offset their soaring AI costs. This has to be one of the worst-kept secrets in corporate America. Even a casual peruser of LinkedIn would have known this a year ago.

Corporate adoption in Axios‘ view is running into four expensive headwinds such as:

• Using AI to automate disliked tasks rather than prioritizing revenue-generating tasks–which is understandable without guidance and pressure on time.
• Using AI for trivial tasks such as checking the weather (well, no one said they couldn’t)
• Leadership is clueless on what AI tools work and are throwing licenses at the employee wall to see what sticks.
• Reluctance to give AI models proprietary information, which makes the AI tool less effective. (Not feeding AI models proprietary information to prevent it from becoming public in LLM models is, one would believe, an understandable concern.)

Even OpenAI’s Sam Altman commented when Anthropic’s Series H was announced that corporate costs are the most valid concern to date.

Unless there is a massive enterprise pullback in AI spend, though, look to Anthropic floating that IPO no later than the fall, even if corporate AI spend pulls back. It’s to be expected. The Gartner Hype Curve is fully in gear and the momentum from Inflated Expectations to the Trough of Disillusionment will continue, until it is processed and moves on to the Slope of Enlightenment.

The speculation is now “official”, since it is by

That train is now approaching, though realistically, no one knows when it is due and at what station.

The need: Oracle must reduce the extent of its “liquidity and capital expenditure crisis” in order to stay in the AI Game. Layoffs of 30,000 staff, or 18% of their global employees, is not enough. A fresh financing of $16 billion from the PIMCO bond fund and others cannot relieve the financial stress created by a previous estimated $72 to $100 billion in previous debt load and payments, so significant that banks refused to lend to still-profitable Oracle. And the AI transformation itself is high risk. Oracle owes OpenAI alone $553 billion in remaining performance obligations, and it has obligations to Meta as well. Add to this the long “taffy pull”–the years-long process of building, chip expenditure, then making a data center operational and generating cash. [TTA 14 May, 7 May, and prior; also Ed Zitron’s article for a much longer take.] Take all of them together, and they are polite words for “rock and a hard place” or a Very Dark Corner.

The London investment banking firm Nelson Advisors has taken a deep yet remarkably easy-to-digest analysis on a potential sale. Highlights are below. The paper is one long web page, not a deck of 50 pages. It is well worth your reading time.

Background: Cerner was bought four years ago in the go-go days of June 2022 for $28 billion. Cerner had an aging EHR and a deteriorating market share. Recently it’s plummeted to a 27% market share versus Epic’s 48% in large health systems. Oracle’s interest was not only in health, but also the health data Cerner contained. The plans were to update the software based EHR to a cloud-native data platform as the linchpin of Healthcare Transformation (Ed. note), except that integration proved to be slow and far more expensive than estimated.

Oracle also inherited from Cerner two huge and impossible to escape Federal obligations: the Military Health System EHR and the Veterans Health Administration EHR Modernization, two separate but mandatorily interoperable systems. MHS was the first implemented and is now  completed, but remains an obligation. The VA EHRM, as TTA has chronicled, started rolling out in 2020 and by 2023 was halted after five implementations Due to Disaster. It resumed in April 2026. The VA and Congressional process for funding now has tight guardrails in place on continuance.  

Who will buy the Oracle/Cerner EHR operation is the question. For how much isn’t as clear. Selling Oracle Cerner “represents the most significant “lump sum” of liquidity available. In the Nelson analogy, Oracle took the Cerner cow, milked it of data to feed its data into its LLMs, and no longer wants knackered ol’ Bessie even rejuvenated by the cloud. (In this Editor’s view, Oracle knows it is fighting a losing battle against Epic, which does privately pretty much what it wants and plans to stay that way.)

The obvious group of potential buyers are ‘hyperscalers’ who view health data as the Next Frontier. They already have feet in this healthcare pond. They also meet approved FedRAMP High security requirements for the VA and MHS contracts. Equally, they all have drawbacks.

Microsoft seems the most logical. It already has a huge footprint and expertise within health systems, courtesy of ambient scribe Nuance/DAX Copilot and cloud computing platform Azure.

• Conflict #1: Epic is a major Azure customer. Would Microsoft be willing to lose this business in a high-stakes move?
• Conflict #2: FTC would likely challenge the acquisition based on this huge existing footprint.

Amazon is also engaged in healthcare, but not with health systems. It has Amazon Health Services comprising Pharmacy, One Medical, and DTC telehealth services. (Editor’s note: not mentioned by Nelson is that Amazon Health has a new leader, Dr. Roy Schoenberg, with experience in Federal contracts via Amwell for the Defense Health Agency and MHS. This broke late last week.)

• Conflict: Amazon Web Services is an established vendor in other areas of health systems, and acquiring an EHR could be seen as too much under one roof.
• Problem: no experience with EHRs (same as Oracle) nor highly regulated health systems. The scale of the MHS/VA implementation and academic hospitals would be a steep learning curve with little existing precedent or credibility in Amazon-World.

Google certainly has the size and resources, and could position the EHR to rival both Microsoft and Epic. 

• Conflict #1: Cultural. Google moves fast and healthcare slowly.
• Conflict #2: Lacks the enterprise sales and support needed to service health systems. It doesn’t have a service culture.
• Editor’s note: Google has tried and failed to be a healthcare giant at least twice. It doesn’t seem to fit.

Nelson also looked at two outliers, UnitedHealth Group/Optum and the hospital groups HCA or CommonSpirit Health. Both would be vertical integrators. Hospital groups do not have the margin nor borrowing power to make the move. UHG and their Optum operation face cash crunches and ongoing Federal scrutiny. (Had this been a few years ago under a different management, this would have been on strategy for UHG.)

Another outlier from the international space is SAP. Their aim would be global expansion into the Middle East and Europe with another asset their enterprise resource planning (ERP) expertise. Their problem? Lack of experience in the highly regulated US environment. In the Nelson view, the US Government could be the make/break for any deal.

The final destination for this ‘hard to sell’ asset? Private equity. And more than one involved. Nelson looked at five PE players in the healthcare space: Thoma Bravo, Francisco Partners, Bain Capital, Blackstone, and New Mountain Capital. (All are familiar PEs to Readers.) Even with their considerable individual assets, it would likely take a consortium to buy Oracle Health in a $20 to $25 billion deal. Nelson rates this as the most likely scenario as long as a consortium could be formed and it can be seen as a turnaround. The drawbacks are a governance structure and the real lack of an exit strategy. (PEs always need exit strategies to keep the funders happy. They are not in it to buy and keep.) The lower price could be made palatable to Oracle if they retained the Oracle Cloud Infrastructure (OCI) network and the Oracle Autonomous Database revenue streams.

The other partner in this consortium scenario? The Federal Government. It’s a high priority to secure the EHR for both the MHS and VA. Congress is already concerned.

Place your bets!  Hat tip to a Reader who wishes to remain anonymous.

Dr. Schoenberg, whom with physician brother Ido co-founded American Well in Boston in 2006, can be considered one of the pioneers of telehealth. Notably, he had become less active in the now Amwell starting in June 2024 when he stepped down as president/CEO, moving to executive vice chairman. His LinkedIn profile has this board role ending this month with the announcement. He is also listed as the founder as of September 2025 of Aileen.ai, a personal health companion targeted to the senior market, and with various board positions that continue. He announced his appointment on a LinkedIn post yesterday.

From reports, Mr. Lindsay led the search for his successor. Lindsay oversaw a restructuring of AHS into six divisions last June, leading to several executive departures. One Medical’s integration and marketing has remained a struggle since 2022 and Amazon Care was discontinued in 2023. The scope of the reorganization and the status of the parts will be a considerable challenge for the good doctor.

Amazon Health Services comprise primary care (One Medical), pharmacy services (Amazon Pharmacy), and telehealth services (Direct Message Care and Video Care). All are tightly integrated with Amazon Prime. (As an Amazon shopper who has studiously avoided the expense of Prime, that has been a drawback.) CNBC, Healthcare IT News

The “Messy Middle” by Molly Kinder in “Kinder Futures”
Why the years between today and post-AGI abundance will be the hardest political and economic problem of our generation, and why we can’t skip past them

This is the longer of the two essays that presents two realities: Reality 1 of today where AI is increasingly everywhere, but the societal effects have yet to be felt, and Reality 3 where it has been absorbed and we are in the bright sunlit uplands of abundance. The mess is in Reality 2, which she posits as year, perhaps decades, of societal upheaval and what we are already seeing–the near-destruction of the professional  ‘cognitive classes’. Yes, people like you and me as creative thought, ability, and specifically cognitive ability, become commodities. This cognitive displacement will affect governments, as the vast middle that pays most taxes suddenly finds income cut in half or more, forced into early retirement, and dependent on government for support. The essential jobs as determined during the Covid pandemic will be the least affected, but still affected because their income trajectory despite need has been declining and that won’t change. The trades survive because their practitioners are scarce for now. The effects on the middle and upper middle classes internationally will be brutal in her scenarios. Ms. Kinder does arrive at conclusions re the political backlash rearranging the current state (we are in it now) and a managed transition that you may not necessarily agree with. See the comments at the end.

The Tech Bros Want to Build God by George Shay in “Common Sense”

Why Pope Leo XIV May Be Right to Worry About Artificial Intelligence—But Not for the Reason You Think

Pope Leo’s first encyclical tackled AI and began, in the introduction, “Humanity, created by God in all its grandeur, is today facing a pivotal choice: either to construct a new Tower of Babel or to build the city in which God and humanity dwell together.” Further, “We must, then, avoid the “Babel syndrome,” namely the idolatry of profit that sacrifices the weak, a uniformity that neutralizes differences, and the pretense that a single language — even a digital one — can translate everything, including the mystery of the person, into data and performance. The risk of dehumanization — of building a future that excludes God and reduces the other to a means — is an ancient and ever-new temptation that today takes on a technical guise.” Like most encyclicals, it is written in exceedingly dense, very Catholic language (disclosure, your Editor is Roman Catholic). 

Mr. Shay’s take is considerably less dense and moves from a different philosophical ground. All too many of the ‘tech bros’, especially at a high level, speak of technology in almost religious terms. AI has become their highest expression of faith. They rejected traditional faith while young and redirected that impulse, that exists in nearly every human save the sociopaths, into technology. To the observant, that is nothing new–witness the well-known lives of Steve Jobs and Bill Gates. Instead, in the search for meaning, their salvation is technology.

FTA: “Instead of God creating man in His image, we now hear man preparing to create a machine in his own image.

Some of the rhetoric coming out of Silicon Valley is astonishingly explicit. Researchers and entrepreneurs casually discuss creating “machine gods” or engineering a “new species” superior to humanity. And importantly, they are not speaking metaphorically. They mean it quite literally.”

The ultimate question is ‘why’–and there is no AI that can answer that, despite how the visionaries position AI “in terms that resemble divinity”. 

Your Editor expressed two comments that are solely her observations. The first: “Perhaps those most enthusiastic about AI are those most frightened of the existence of God behind life, faith in that God, ethics beyond their own wants and needs, consideration of their fellow man, and their own mortality. Is it no mistake that most of those leaders in technology resemble flawed and strange little boys, even when they are 70? That is an unsettling thought.” The second is lengthy and saved for your reading of the article.

Garner Health’s doctor finder app now has another $100 million to back it. The Series E was led by Index Ventures with participation from existing investors including Kleiner Perkins, Redpoint, Thrive, Sequoia, Founders Fund, and Kaiser Permanente Ventures. The round brings Garner’s valuation to $2.74 billion. It’s a far cry from the 5x Series E typical raises of a few years ago, but fairly healthy in the current slow recovery of lettered raises and more conservative fundings. Previous funding was $219 million through a 2024 Series D. Garner’s apps (Assistant and Research Agent) are targeted to employers and benefit managers as a service to connect employees with high-performing in-network physicians, based on clinical performance data and claims information, and to save on total healthcare spend. The company claims that they serve over 700 companies. The app is still available to individuals as a free download. Release 

Veradigm finally files its financial reports for 2023 and 2024, and they’re not cheering. This took the form of a ‘Super 10-K’, a comprehensive annual report that covers two fiscal years, as promised on Veradigm’s investor call last February. 

2023 to 2024 results weren’t cheering for the shareholders, as revenue dropped 5% 2024 compared to 2023, continuing their multi-year decline.  The company in 2024 swung to a net loss, a change of $341 million due to impairments and other costs. Adjusted EBITDA was affected by 29%.  

• Revenue of $594 million in 2024 compared to revenue of $626 million in 2023, which includes $19 million of non-recurring revenue in 2023 from settlements.
• Net Loss in 2024 was $292 million compared to Net Income of $49 million in 2023. The net loss in 2024 reflects an impairment of goodwill and other assets as well as the impact from transaction and other costs.
• Adjusted EBITDA of $94 million in 2024 compared to Adjusted EBITDA of $132 million in 2023.

Veradigm has three main market segments: provider, payer, and life science. It is adjusting its software and implementations to incorporate AI, which was briefly discussed on their earnings summary call 26 May. The transcript and audio is on Seeking Alpha.

According to Veradigm’s release, they will file their 2025 Form 10-K before year-end 2026. Revenue is estimated to be in a range between $584 million and $589 million, flat from 2024. Nothing definite was stated about their 2026 filings with the SEC, which will be required before they can file for relisting on Nasdaq or on any exchange. From the release, “The Company remains focused on getting current and staying current with its SEC filing obligations and subsequently relisting its common stock on a national exchange. As of May 12, 2026, the Company had 109.0 million shares of common stock and 13.3 million of unvested restricted stock units outstanding.”

Veradigm also has a new CFO, Christian Greyenbuhl, replacing interim CFO Lee Westerfield, who came for six months and stayed for two years. He will remain in a ‘strategic advisory role’ until March 2027, a standard workout period. Mr. Greyenbuhl was formerly CFO of Ministry Brands, which provides over 90,000 faith-based organizations and private sector businesses, SaaS payments and background screening solutions. Previously, Christian was a longtime finance and operating executive at ADP. 

It’s now the fourth year of make-up reporting since 2022. Veradigm’s their failure to provide audited reports due to financial software problems has been an embarrassment for a software and once-leading health IT company. In 2025, 15% of staff were cut, three facilities closed, and will exit at least two more as well as their present Chicago HQ by 2027. Our February article also provides a capsule of Veradigm’s recent troubled history and our back file as the situation for the former Allscripts developed.

And to wind up with something that could really help in hospitals in moving patients from point A to B…Rovex has developed a robotic transporter for in-hospital patients. The transporter, Rovi, is the world’s first autonomous transport robot that can transport any hospital stretcher yet maintain contact with the care team. It is being piloted without patients at BayCare Health System’s Morton Plant Hospital. The pilot is designed to evaluate workflows, transport patterns, staff burden effects, and operational opportunities within the hospital environment. Later phases will support potential in-hospital robotic stretcher movement. Those of us with hospital experience either in the clinical setting or as patients know how speed of transfer affects the entire clinical workflow and care. BayCare is the largest not-for-profit academic health care system in West Central Florida. Rovex is a two year old company, still in seed funding of $2.27 million (Pitchbook) and headquartered in Gainesville, Florida.  BayCare release, Interesting Engineering

From KLAS’ brief unpaywalled summary:

• The number of hospitals impacted by EHR purchase decisions dropped 40% compared to 2024 and nearly 50% compared to 2023.
• Acute care EHR purchase energy declined significantly in 2025, driven by several converging factors. Ongoing questions around government policy contributed to hesitation, and at the same time, many health systems redirected investment toward technologies with more immediate financial returns, such as AI and other solutions designed to improve operational efficiency.
• Additionally, while Oracle Health has continued to lose market share, many Oracle Health customers have been deferring action as they await greater clarity on the vendor’s evolving direction.

In this stagnant picture, the commenter ‘Spinout’, who clearly has report access, adds that:

• Epic continued to take customers away from Oracle Health, adding 77 hospitals with 19,000 beds in 2025 while Oracle Health lost 56 hospitals with 15,000 beds. (This is absolute confirmation of what this Editor has heard from industry people–that Oracle is now an also-ran EHR.)
• Oracle Health customer satisfaction continues to go down. One-third of its customers say Oracle isn’t in their long-term plans while another one-third say they might want to leave or would if they could. That leaves the company’s AI-powered EHR as a make-or-break product. (When you can’t hold 2/3 of your customers, you’re in big trouble. And an AI-powered EHR? With a shortage of AI data center capacity, TTA 14 May, welcome to downtime)
• Third-place vendor Meditech continues to increase customer retention as its legacy clients increasingly choose Expanse instead of a new vendor.

What this may mean. It’s becoming all too easy, too logical, to make the case that Oracle Health has turned into a losing proposition, one that Larry Ellison can no longer afford even if he owns 40% of the company. One senses a pile of dirt, a hole, and a shovel. Mr. Ellison also needs money, which makes it even more likely that Oracle Health is spun off or sold for cash. 

From this Editor’s earlier article about the 30,000+ Oracle layoffs (10,000 in the US, with 539 employees laid off effective 26 May-1 June in Kansas City, Cerner’s former HQ and presumably where most EHR staff remained): Is Health, once the focus of Oracle’s Big Transformation, now just a used and broken toy? What’s the future of Oracle Health if the strategy is AI 24/7 and EHRs and healthcare system SaaS just do not fit the picture anymore?

Yet Oracle cannot simply close the division and turn off the systems. In addition to hospitals and health systems, let’s not forget that the VA is back to sprinting towards an implementation schedule ending in 2031. The EHR Modernization has 13 VA locations to cutover this calendar year, with 26 new sites in 2027 and 28 in 2028. Oracle will literally jeopardize every Federal contract they have if they botch VA EHRM. In short–Oracle Health is one of several bottomless pits Oracle faces.

Walgreens’ leases on the market–see A&G Real Estate Partners. One of the outcomes of Walgreens’ reorganization is a drastic downsizing of its retail footprint. Both city and suburban/rural locations have been affected, to the distress of many communities. Of note is the organization that the New Walgreens has chosen in this process is A&G Real Estate Partners, which specializes in companies exiting their retail locations due to Chapter 11 bankruptcies, such as Saks Global, Party City, over 1,000 Rite Aid leases, and Bed Bath and Beyond. A&G is marketing a mix of 78 Walgreens store leases and owned properties nationwide, per their presence at the ICSC retail real estate confab in Las Vegas. Sycamore Partners, the new owner, has deep experience in retail, so that their choice of real estate marketer is not surprising. One can expect that number to increase as the three-year plan has 1,200 closures and Walgreens, in layoffs announced earlier this year, reiterated that they were closing dozens of locations this year [TTA 27 Feb]. FTA, “A&G expects the same kinds of retailers that were interested in the Rite Aid stores to be interested in Walgreens, as well as possibly specialty grocery chains and healthcare providers, according to (principal) Joe McKeska.” A&G principals at ICSC met with these retailers as well as lenders and investors. CoStar

On to cheerier news

Oura has confidentially filed an S-1 with the Securities and Exchange Commission (SEC) for an IPO. The Finland-based smart ring developer did not state pricing nor number of shares in the preliminary filing.

Pre-IPO, Oura is on a roll as well, with up to 5.5 million devices sold, $1 billion in revenue, and 1,200 health partners. Their investments have been substantial. Oura’s most recent jumbo Series E of $900 million in October 2025 topped a Series D of $250 million in 2024, sandwiching a debt financing of $250 million in September 2025. Major investors are Fidelity and Dexcom, the latter announcing a partnership and $75 million investment in November 2024. Oura has also been on an acquisition tear: Doublepoint in March, Sparta Science and Veri in 2024, and Proxy in 2023.  Last October, FDA authorized a trial with Oura users of a new blood pressure feature. It also has partnerships in women’s health and hormone tracking. Most recently, last week Oura announced a partnership with Resmed, giving users access to a sleep assessment and educational resources.  Oura release, Healthcare Dive, Mobihealthnews

Swoop just scooped up NimbleRx. Acquisition cost was not disclosed. Swoop is a data platform for life sciences companies that enables marketing to both patients and healthcare providers, engaging them via targeting, community engagement, AI-powered web solutions, coordinated omnichannel activation, and now prescription fulfillment. NimbleRx streamlines workflows for pharmacy providers and prescription management for both pharmacies and patients, and claims coverage of 16 million patients. Their service will now be known as SwoopRx by Nimble. This is Swoop’s second acquisition in the past year, with MyHealthTeam, an opt-in patient social network for 62 conditions. Release, Mobihealthnews

A slower news week preceding the Memorial Day holiday in the US and the UK late May bank holiday.

Federal budgets for 2027 are in the Congressional washing machine, and the cycle is on ‘agitate’. VA Secretary Doug Collins has tagged a 25% increase in the EHR Modernization budget for FY 2027 over what is currently in the 2027 Military Construction and Veterans Affairs Appropriations bill –$4.2 billion versus $3.4 billion, an increase of $840 million. He testified on Wednesday 20 May to the Senate Veterans’ Affairs Committee and Thursday 21 May to the House Appropriations Subcommittee on Military Construction, Veterans Affairs and Related. Apparently, the biggest problem VA has with the much-repaired and now standardized Oracle EHR is that every VA executive director wants it now, not later. An additionally funded EHRM would speed up the cutover for VA facilities to go from ‘dial-up’ to ‘cyberspeed’ internally, in communicating with other VA hospitals, community care, and in record sharing with the military system and civilian health facilities.

Difficulties reported to date (April for four sites in Michigan, VISN 10) are around transferring health records between VA and Department of War facilities. DoW healthcare also uses Oracle, but a different version suited for their needs that has been fully implemented. 

While the House has already passed the bill at the lower budget number and sent it to the Senate, the subcommittee chair John Carter (R-Texas) during the hearing said they’re “not through with the possibility of getting you some more money”. 

VA’s implementation timeline is 19 before the end of this year (13 new and the 2020-24 six), 26 new sites in 2027 and 28 VA Medical Centers in 2028. Even sped up, there are still 90 more to go and the deployment is not expected to be complete till 2031. FedScoop 21 May, 30 April

Update on the fraud indictment of the former EHRM director, John Windom. Surprisingly, there has been little to no mainstream media coverage of the Federal charges against John Windom, who was indicted on 25 March in the Federal District Court for the District of Columbia. The three counts related to accepting cash and gifts from vendors plus failure to report them could bring a maximum of 35 years. This article on conservative news website PJ Media is the most recent (re)telling of the tale and links to nearly all the same sources this Editor included in our 3 April article. It is more colorful than our reporting but brings up an important point I overlooked: where, oh where, are the indictments of some of the vendors who doled out cash, gifts, and maybe more, and in return got prime and sub-contracts. He knew, they knew to keep quiet–‘loose lips sink ships’. Because any Federal contractor–I worked for two, Viterion Digital Health and Collaborative Health Systems, then part of WellCare Health Plans–receives compliance training on working with their Federal agency counterparts. 

Perhaps there are investigations and indictments to come, as I’ve seen in Federal Medicare fraud cases that peel like an endless onion over years. According to the VA inspector general, Mike Missal, who served from 2016 until January 2025, evidence was being gathered internally back during the Biden administration. This fits the timeline of the US Attorney requesting a grand jury be summoned then sworn in on 30 October 2025. Mr. Missal was fired along with 16 other inspectors general by the incoming Trump administration.

Since Mr. Windom was deeply engaged in the choice of Cerner for the VA EHR in 2017-2018, and in the disastrous implementation of VA Mann-Grandstaff (VISN 20) in October 2020 and four more in 2022, resulting in the rollout’s termination in 2023, Oracle would be unwise to not prepare for a few questions about Cerner’s relationship with Mr. Windom, as I wrote at the time. 

The PJ Media article also references the comprehensive article in the 27 March Spokane Spokesman-Review, which has been on the Cerner/Oracle implementation story since the implementation failure in the region’s Mann-Grandstaff VA facility. Their check of the OEHRM website as of that date confirmed that Mr. Windom was still listed as the deputy director of the Federal Electronic Health Management Office, the joint VA-DOD initiative in the role he assumed in January 2022 after the Mann-Grandstaff problems detonated and the then-Secretary reorganized the department. (Heads did not roll, but they rarely do with SES members). FTA: “The Federal Electronic Health Record Modernization Office did not respond on Thursday (26 March) when asked if Windom remains employed there.” The article by Orion Donovan Smith is a recommended read.

In the funding/M&A department

Healthcare software integrator Commure received a $70 million funding from current investors. Commure’s lead investor is General Catalyst. Commure now has $750 million raised and a $7 billion post-money valuation for its AI infrastructure development. Its subsidiary, Athelas, provides AI-based revenue cycle management and clinical workflow tools. The General Catalyst funding of $200 million plus is an interesting scheme, in that GC fronts the cost of sales and marketing and, in return, receives a share of the revenue from new customers generated by that investment, up to a fixed cap. The new funding will be used for scaling its RCM and practice management platforms, advancing the ‘shared intelligence layer’ beneath Commure’s workflows, and expanding their AI infrastructure into global healthcare markets. Release, Mobihealthnews

Innovaccer acquires CaduceusHealth, a revenue cycle management (RCM) and management services (MSO) provider. Neither transaction cost nor management transitions were disclosed. Well-funded Innovaccer ($675 million through a Series F) has been growing in AI-centric healthcare IT services mainly through acquisition. CaduceusHealth is the fifth in their creating a “comprehensive agentic stack” for health systems and provider groups in their Flow suite. Innovaccer claims to serve over 200 health systems and payers, 95% of community pharmacies, and 80 million patient lives across the US. Release Unfortunately, their growth has been matched by a reduction in staff, with 340 layoffs in the US and India. It is their third layoff in four years as it applies its own AI to automate its own processes. (We are seeing a lot of this across the board, allegedly.) FierceHealthcare

We close with a major Must Read with the OpenEvidence-Doximity battle.

OpenEvidence and Doximity are slugging it out for the same market funding–and a third competitor has just sneaked into the ring. OpenEvidence is the upstart, founded four years ago, and the best valued ($12 billion) yet private healthcare AI company on the planet Earth and is generally thought of as the up-and-coming platform for physician information. Doximity is the mature company, public with a $3.6 billion market cap, proven revenue of $645 million, and (be still my heart) profitable with an EBIDTA margin of 55% and a stunning 49% free cash flow margin. It’s been dubbed ‘LinkedIn for doctors’ but is actually much more with tools for secure telehealth, news, reputation management, and free CME.

They are mutually litigious. Both OpenEvidence (OE) and Doximity tag-team each other in product offerings, use defamation tactics and key staff poaching, and in product development, copycat each other, with Doximity generally leading development and OE following shortly thereafter. Coming up is Doximity’s new product, an in-workflow e-prescribing, prosaically called Doximity Prescribe. Based on the pattern, how long will it be before OE develops a similar product?

Where they make their money is only indirectly from users. Both are supported by a fixed source–pharmaceutical advertising. They both slug it out for physician attention. While doctors love (or hate) both, if they become too similar, the balance will tip. Into this bout steps OpenAI with a new professional product, ChatGPT for Clinicians [TTA 30 April]. Lurking near the ropes is the AI-powered iteration of Wolters Kluwer’s UpToDate peer-reviewed medical content, integrated with Microsoft and Abridge, already in 70% of the largest enterprise health systems because it’s been around forever. OE’s vulnerability may be overpromising in claiming ‘no hallucinations’ of their AI-generated medical content–a claim that is structurally impossible, and results in deficits in completeness, communication quality, and systems-based safety reasoning.

Digging through all of this is the intrepid Sergei Polevikov on his Substack AI Health Uncut. Grab a cuppa and sandwich for this one. For most of the article (Part 1 of 2!), a subscription is required. Consider it money well spent for access to some of the best investigative reporting around with plenty of backup. OpenEvidence Prescribe Coming to Your Doctor’s Office This Month?

TTA has an open invitation to industry leaders to contribute to our Perspectives non-promotional opinion and thought leadership area. Today’s topic is about the use of AI in the mental health area–how it is uniquely exposed to risk from LLMs and generative AI–and where best to use them. The author, Hari Prasad, is co-founder and CEO of  Yosi Health, a full-service technology ecosystem that connects patients with their providers through the entire care journey before, during and after the visit, modernizing the

One in three U.S. adults has now used an AI chatbot for health information in the past year, according to the most recent KFF tracking poll. Among adolescents and young adults, one in six has turned to a large language model specifically for mental health advice. That second number should stop every behavioral health leader in their tracks, because when those tools get it wrong, the consequences are not a bad product recommendation. They are clinical.

We are not witnessing patients casually Googling symptoms anymore. They are having extended, emotionally charged conversations with generative AI tools about medications, crisis management, and care decisions, often weeks or months before they ever speak to a licensed provider. And the technology they are confiding in has a well-documented failure mode that the industry still has not adequately addressed: the hallucination.

The Problem Is Not That AI Sounds Wrong. It Is That It Sounds Right.

An AI hallucination occurs when a model generates a response that is fluent, confident, and completely fabricated. In most consumer contexts, that is an inconvenience. In behavioral health, it is a patient safety event waiting to happen.

Consider the real scenarios already playing out: a generative AI tool confidently suggesting an incorrect dosage for a mood stabilizer. A chatbot offering therapeutic advice that directly contradicts evidence-based protocols for trauma. A model providing reassurance to a patient in acute crisis when the clinically appropriate response is immediate escalation. These are not hypothetical edge cases. They are the predictable output of systems designed to be helpful and conversational above all else.

What makes this uniquely dangerous is the packaging. These models are engineered to sound empathetic and authoritative simultaneously. A patient in a vulnerable mental state has no reliable way to distinguish a clinical fact from a statistically plausible guess. And once that trust is broken, the damage extends beyond the misinformation itself. A patient who receives hallucinated guidance during a moment of crisis may lose willingness to engage with the legitimate care system at all.

Why Behavioral Health Is Uniquely Exposed

Two structural realities make behavioral health the highest-stakes frontier for AI hallucination risk: subjectivity and scarcity.

First, subjectivity. Unlike a fracture that shows on an X-ray, mental health concerns are communicated through nuance, context, and subtext. Generative AI is exceptional at mimicking tone. It is incapable of the clinical judgment required to understand the weight behind a patient’s words. There is no lab value to cross-reference, no imaging to confirm. The entire diagnostic framework depends on the kind of human interpretive skill that cannot be approximated by next-token prediction.

Second, scarcity. As of late 2025, 137 million Americans live in a Mental Health Professional Shortage Area, and only about 27% of need is being met in those regions. Appointment wait times range from three weeks to six months depending on location and specialty. For the roughly 29.5 million adults with a mental health condition who received no treatment last year, the barrier was not awareness. It was access. When the pathway to a licensed professional is blocked by a three-month wait, an AI chatbot becomes the path of least resistance, not by preference, but by default.

This is the uncomfortable truth the industry must confront: patients are not turning to AI because they trust it more than their providers. They are turning to AI because the system has not given them a better option.

The Fix Is Infrastructure, Not Disclaimers

The standard industry response to AI hallucination risk has been to add disclaimers and guardrails to the models themselves. That approach treats the symptom while ignoring the disease. The reason patients are having clinical conversations with chatbots is that the operational infrastructure of most practices still makes it easier to talk to a machine at 2 AM than to get a timely appointment with a human being.

The real solution is to shift AI’s role from clinical logic to operational logistics. Practices need to deploy technology that removes the administrative barriers driving patients toward unregulated alternatives in the first place. That means rethinking three operational layers.

Deterministic intake over conversational intake. When an LLM “chats” its way through a patient intake, it introduces the same hallucination risk we are trying to eliminate. The alternative is structured, deterministic intake systems that gather discrete clinical data without improvising questions or advice. Symptoms, history, social determinants of health, all captured through validated frameworks and delivered into the EHR as clean, fact-based data. The clinician gets a head start. The patient gets accuracy.

Precision navigation over generic triage. AI’s greatest operational strength is processing complex variables at scale. That capability should be pointed at routing, not counseling. If a patient’s digital intake surfaces indicators of acute risk, the system should not offer a supportive quote. It should trigger immediate escalation to a crisis line or emergency clinician. The technology’s job is to get the right patient to the right level of care at the right time, not to play therapist in the interim.

Intelligent follow-up over passive waiting. The period between appointments is where behavioral health patients are most isolated and most vulnerable. This is where AI can add genuine value, not by providing care, but by acting as a monitoring layer. Structured check-ins, flagging of concerning patterns in patient-reported outcomes, and automated alerts to clinical teams when intervention thresholds are crossed. The AI serves as a tripwire, not a therapist.

From Advice to Access: The Shift That Matters

The practices getting behavioral health engagement right are not the ones deploying the most sophisticated AI interfaces. They are the ones using technology to collapse the administrative distance between a patient’s first expression of need and their first clinical encounter. When scheduling, intake, and insurance verification happen before the patient walks in, the clinical encounter starts on solid ground. That is what patient engagement actually looks like in 2026.

The hallucination problem is not an argument against AI in healthcare. It is an argument for precision about where AI belongs. Every time we deploy AI in the clinical layer without adequate safeguards, we introduce risk. Every time we deploy it in the operational layer to accelerate access, we reduce risk. The distinction is not subtle, and the stakes are too high to keep blurring it.

Behavioral health already has a trust deficit driven by stigma, scarcity, and systemic friction. The last thing this field needs is a technology layer that erodes trust further by giving patients confident answers that turn out to be wrong. The opportunity in front of us is to use AI to rebuild that trust by making the system itself faster, smarter, and more responsive. Not by replacing the clinician, but by making sure the patient actually gets to one.

Related article from TTA

Character.AI sued by Pennsylvania on its chatbots posing as licensed physicians and psychiatrists