How the Trump administration is recasting government’s role in regulating health technology

STAT News – Tuesday, December 23, 2025
By Casey Ross, Mario Aguilar, Katie Palmer, and Brittany Trang

In the technology industry, there is a proud tradition of unleashing innovations before regulators put the brakes on progress to consider concerns about safety and fairness.

But in digital health under President Trump, it’s the regulators who are endorsing — even accelerating — the pace of change.

During the past year, Trump’s health agencies swiftly adopted artificial intelligence and other technology products and dismantled guardrails designed to protect patients. Policy experts said the combination of rapid uptake and reduced regulation requires Americans to accept the dangers of unchecked experimentation to confront global threats to the nation’s leadership in AI.

“That’s, in effect, what’s being asked in the health care sector, which is that we set aside our concerns about potential risk to individuals and individual patients in the battle for dominance of the world,” said David Blumenthal, professor of the practice of public health and health policy at the Harvard T.H. Chan School of Public Health.

The Trump administration’s approach stands in stark contrast to the one pursued under former President Joe Biden, who called out the particular threats posed by AI’s use in health care in an executive order that proposed a cautious approach. Trump swiftly rescinded that order and replaced it with a strategy that places greater emphasis on supporting innovation and economic progress. The president’s health appointees have sought to undermine efforts to develop AI standards by a coalition of industry and academic partners. And earlier this month, Trump issued an executive order aimed at blocking states from crafting or enforcing their own rules around artificial intelligence, citing the need to remain competitive with China.

All of this is occurring at a time when AI is advancing rapidly, with technology companies unleashing a wave of products promising to automate everything from the documentation of patient-doctor conversations to disease diagnosis and the interpretation of medical images.

Meanwhile, Trump’s health leaders have also proposed to incorporate a broader group of digital health tools to shake up the care of millions of Americans, and sought to create a freer flow of health data to support care and product development.

Taken together, these policies seek to transform the delivery of medical services and streamline the evaluation of a broad swath of health products by government agencies. Here is a closer look at the changes to watch in the year ahead.

Modernizing Medicare with technology

One of the clearest examples of the Trump administration’s proactive approach to health tech adoption is unfolding in Medicare.

As the year wound down, the Centers for Medicare and Medicaid Services unveiled a 10-year experiment to test whether digital tools can help older patients manage their diabetes, high blood pressure, depression, or other chronic illnesses.

Trump administration officials are bullish that the pilot program, called ACCESS, will remove barriers to the adoption of new tech tools, and help primary care doctors incorporate them into their practices.

The goal is to encourage adoption of the technology by changing how doctors get paid, Jacob Shiff, chief AI and technology officer at the CMS Innovation Center, said during a Washington, D.C., health AI summit shortly after the program was announced. Instead of reimbursing every individual service to manage chronic illnesses, as Medicare does now, the new pilot will make recurring payments to doctors for improving patients’ health with the use of technology.

Shiff said the program “gives the clinical organization far more flexibility around how they can achieve those outcomes.”

During the same event, deputy health secretary Jim O’Neill suggested the program could help drive the adoption of artificial intelligence tools that aren’t currently reimbursed by insurers. “I do really see it as a first step,” he said. “At some point I think all payers, including private, will see the advantage in acknowledging that AI and tech are important parts of efficient, high-value medical care.”

The biggest question is how the administration will assess the performance of participants, said Jennifer Goldsack, chief executive of the Digital Medicine Society, a trade group for digital health companies. That will require defining clear improvement benchmarks for patients with diabetes, obesity, kidney disease, and other conditions that commonly afflict older adults.

“We need to have the right measures, but it’s not just the measures,” Goldsack said, adding that health officials must also set clear goals for the amount of improvement digital health products must bring about to be considered successful.

At the sites testing new digital health tools, 10% of patients will be given standard care to help establish the effects of the technologies. But so far, the administration has not specified how data on health outcomes will be analyzed and reported — details Goldsack said are crucial to the program’s success.

Sarah Meier, director of payment and care delivery policy at Mayo Clinic, like many policy watchers, is keenly aware that in a fast-moving administration models may evolve quickly.

“ACCESS is starting in one place, that doesn’t mean it’s where it’s going to stay,” she said during a December policy summit hosted by the American Telemedicine Association. She said where ordinary policy cycles might take five years, it’s not unreasonable to expect that the administration will make changes next year. “I think that’s really important to note the dynamic environment we’re living in right now.”

Automating waste reduction

While the Trump administration is leaning on technology to improve care, it’s also using it to cut waste.

Starting in January, six states — New Jersey, Ohio, Oklahoma, Texas, Arizona, and Washington — will use AI to determine whether Medicare should pay for certain medical procedures requested by patients and their doctors. The treatments due to get extra scrutiny under the prior-authorization pilot include things like skin substitutes and electrical nerve stimulation, where the evidence of medical benefit is shaky.

“They stayed in that space where it’s almost more of an anti-fraud measure,” said Christina Silcox, research director for digital health at the Duke-Margolis Institute for Health Policy. It remains to be seen whether the program will eventually be expanded to include additional services and treatments, akin to the level of review imposed by private insurers.

Tech vendors that participate in the program, dubbed WISeR, will be paid a percentage of savings attributed to their review process; payments will be adjusted downward if the vendor does not hit quality targets.

But important details remain fuzzy. Government documents do not explain how the quality targets chosen will actually translate into payment. In the past, said David Introcaso, a health policy consultant, CMS would provide examples to help stakeholders understand how payments will be calculated, but such details have not yet been publicly disclosed.

“How are they going to evaluate this [program] if nobody knows the criteria they’re using to make these decisions?” he asked. “And from a financial side, nobody knows the formula for paying these contractors.”

A CMS spokesperson told STAT that the agency “has an independent evaluation contractor and is planning to release public reports on WISeR annually starting in 2027,” but did not answer questions regarding who will be doing program evaluation.

Freeing health data

Trump’s health department has taken several steps to free up health data so it’s easier for patients, providers, and health tech companies to use it to make care better and more efficient. And the administration has chosen to encourage the changes it wants by nudging industry rather than wielding the forceful arm of regulation.

Ryan Howells, a principal at Leavitt Partners and longtime evangelist for modernizing health IT, said that the Trump administration’s initiatives represent the most significant progress since the government subsidized adoption of electronic medical records during the Obama administration.

“The reason we’re so giddy about the fact that this administration has prioritized this so much is just the fact that all this stuff has had to happen forever,” he said.

Howells and others have for years been working to promote the easier exchange of health data to ease the burden on patients who often struggle to access medical history that’s siloed across difficult to use portals. Better access to data could speed up new health tech development and facilitate proactive care and better coordination between providers.

Traditionally, the health department’s tech regulator, the Office of the National Coordinator for Health Information Technology (ONC), has led with rulemaking efforts intended to realize this goal. And under the Biden administration, this office was expanded to become the Assistant Secretary for Technology Policy (ASTP).

Under President Trump, by contrast, CMS has taken the lead on an industry-focused approach that leans away from regulation — but still has the support of ASTP/ONC. Early on the two agencies issued an expansive request for information on roadblocks to health tech innovation and adoption, reflecting the tech expertise hired into the administration, including the director of the Center for Medicare Chris Klomp, adviser Amy Gleason and others who were previously entrepreneurs.

From there, CMS moved quickly. In July it made a big show of announcing dozens of health organizations, health IT companies, big tech companies, and startups that had signed pledges to work together to advance various goals, including creating apps to make it easier for patients to access data or on creating new “CMS Aligned Networks” that will enable streamlined access to patient data.

Whether the spirit of cooperation — and good publicity — from the voluntary pledges translates to tangible and lasting progress without rules to back them up will become clearer as the initial deadlines approach next year.

Elsewhere, the health department has vowed to start enforcing rules that forbid health systems or electronic medical record companies from preventing patients, authorized providers, and vendors from obtaining patient data for legitimate purposes. To that end, the Health and Human Services Office of Inspector General recently posted a job listing for attorneys to focus on these investigations.

In the past HHS has not had the bandwidth to investigate complaints of information blocking — an ASTP dashboard currently lists 1,400 claims — but perhaps the threat of legal action will be enough to incentivize most to get in line.

Unleashing AI

While the Biden administration focused on erecting guardrails for AI, Trump has so far chosen to let it run free, especially within federal health agencies.

Peter Pitts, a former Food and Drug Administration associate commissioner and president of the Center for Medicine in the Public Interest, said that quick adoption of AI at the FDA shows that the administration “recognizes the importance of becoming competent in addressing it — not just in addressing sponsor’s uses of it, but in their own uses of it.”

Within the agency, leaders swiftly rolled out “Elsa,” a generative AI model agency staffers can use to speed up regulatory reviews. But some employees characterized the launch as rushed in interviews with STAT and expressed concern that agency leaders both overinflated its capabilities and failed to set clear ground rules. Just months later, the regulator doubled down with “agentic AI” tools to support premarket reviews, and administrative tasks.

Policy and safety experts have expressed concern that the broad deregulatory approach will paradoxically harm technological progress. “At the front lines of care, regulation can be a boon to use,” said Blumenthal, the Harvard health policy professor. “And I think that’s something that the administration has not come to grips with.”

Some developers of generative AI tools are clamoring for a clearer directive from the FDA on whether products they might spend millions of dollars developing will be subject to regulation.

“I would love to have some more formal statements, whether that’s guidance, whether that’s FDA stating [its position] publicly at a hearing or a meeting — even a tweet at this point,” said Katherine Goodman, who develops and researches generative AI tools at the University of Maryland.

So far, Trump has been clear he doesn’t want others jumping into the federal regulatory void with patchwork rules that could hamper progress. A new executive order, sure to be challenged in court, blocks states from enforcing AI laws or crafting new ones.

That order only places more pressure on the federal government to set forth clear rules. “It would be helpful both for AI developers and health care institutions to have a centralized regulatory legislation and expectations, so we know what the bar will be,” said Danielle Bitterman, an oncologist at the Mass General Brigham health system in Boston who builds AI tools. “Developers need that to be incentivized to continue innovating.”

Meanwhile, the FDA’s expertise in both internal and external technologies has been challenged repeatedly this year.

In February, layoffs hit AI experts in the agency’s device center especially hard, including those in its research division. The Office of Digital Transformation, responsible for internal rollouts of information technology to support the FDA’s regulatory work, was also impacted.

Some of those roles were later reinstated, but some staffers had already initiated new job searches and decided to take their expertise elsewhere. “What happened during the early days of the Elon Musk RIF era was just thoughtless execution of foolish policy,” said Pitts. “The FDA is paying the price for that now in dozens of ways. Least of which is being properly staffed, not just to implement a digital strategy, but to build a digital strategy.”

The FDA has shown some signs of life indicating it’s paying attention to the challenges posed by artificial intelligence. In November, it held a meeting of the Digital Health Advisory Committee to discuss how regulation might be applied to different kinds of generative AI-enabled mental health products, like therapy chatbots. The agency also created a cross-agency AI policy council to weigh issues of regulation. People with knowledge of the situation say they believe it may be possible for the FDA to update guidance documents to provide clarity to AI developers without running afoul of Trump administration directives to pull back on regulation.

As the FDA sees dramatic turnover in senior leadership across its offices, key digital health and device leaders have also left the FDA throughout the year.

In the absence of consistent leadership and expertise, some health policy experts are concerned that the administration’s move-fast-and-break-things approach to health technology adoption will lead to lax regulation that endangers patients.

Biden’s FDA under Robert Califf spent a good amount of time talking about the regulation of generative AI and issuing guidance on the use of AI in drug regulatory submissions and in constantly-evolving algorithmic medical devices. “I don’t know what the current administration’s position is,” said Pitts. “I hope it’s not, ‘Let us just close our eyes and jam our foot down the accelerator and see what happens.’”

Access this article at its original source.