AI health apps have been given an unusually direct route into US healthcare policy under the Trump administration, with technology companies invited to help shape regulation through a private Slack workspace and promised access to millions of Medicare users.
The year-long campaign has involved senior officials from the Department of Health and Human Services and the Centers for Medicare & Medicaid Services, alongside executives from companies including Microsoft, Anthropic, OpenAI, Apple and Google.
In February, the Food and Drug Administration invited at least 35 industry organisations to a closed meeting about conversational AI tools, including patient chatbots. The session was not listed on the FDA’s public calendar or regulatory notices, and the invitation was circulated through Slack rather than through a public process.
Companies were asked to explain how their products worked and how they accessed patients’ sensitive medical information. Morgan Taylor, a senior policy adviser at CMS, described the meeting as an opportunity to help shape future guidance on the “safe Conversational AI development and monitoring”.
Thousands of Slack messages, meeting transcripts and video recordings reviewed by KFF Health News show technology executives communicating directly with senior officials, including Amy Gleason, the former acting administrator of Elon Musk’s Department of Government Efficiency.
The discussions formed part of the administration’s wider “Health Technology Ecosystem” initiative, which aims to make medical records more accessible and integrate digital tools into healthcare. Officials have also been developing schemes that would allow companies to bill Medicare for services such as wearable-device monitoring and AI chatbot advice.
Private technology group under scrutiny
The Slack workspace was established in August 2025 and grew to about 1,700 members. Participants included representatives of digital wellness firm Oura Health, AI company Palantir and investment firm 8VC, while relatively few doctors, hospital representatives or patient advocates were involved.
A code of conduct circulated by Ms Gleason said the group was not an “advisory committee” and would not be used to obtain advice or recommendations for HHS or CMS officials. It also stated that federal employees were prohibited from endorsing products, services or businesses.
Joseph Daval, a former FDA lawyer and research specialist at Harvard Medical School, said the group nevertheless resembled a federal advisory committee because members interacted and built consensus around possible courses of action.
Formal advisory committees are subject to requirements intended to ensure independence and a fair balance of interests. Mr Daval said those safeguards were designed to prevent industry groups from capturing government policy.
CMS officials have not answered questions about the legality of the Slack workspace. In a statement, Ms Gleason said the initiative was an “open, voluntary technical collaboration” intended to improve access to health information and useful digital tools.
The FDA did not respond to a request for comment.
Medicare app library described as a “sales engine”
During online meetings, CMS officials told technology leaders that the government would help promote their products. Jacob Shiff, the agency’s chief AI and technology officer, said in February that he hoped the programme would become a “sales engine” for health apps.
“The people on this call who hopefully will have amazing solutions that work will be the winners,” Mr Shiff said. “So hopefully we’re selling — or enabling you guys to thrive.”
A recording of the meeting, which had been publicly available, was made private after questions were raised about the comments.
CMS later launched a Medicare App Library featuring about two dozen commercial applications aimed at older and disabled Americans. The products offer tools relating to conditions including obesity, cancer and diabetes, and some say they can analyse health information or develop plans for users.
At an April meeting, Mr Taylor suggested that appearing on the Medicare website would encourage patients to trust the products. CMS officials have also indicated that apps listed in the library could be prioritised for a programme reimbursing companies for AI advice and wearable-device tracking.
One of the apps listed, Slothwise, was launched this year and charges users 9.99 dollars a month to analyse medical records and other health data. Its chief executive, Sofia Sigal-Passeck, said the product was currently informational and had been built by doctors, researchers and engineers.
Doctors have raised concerns about the limited evidence surrounding some of the technology and the absence of clear liability when tools provide inaccurate advice.
“Some of these tools just aren’t ready for primetime,” said John Whyte, chief executive of the American Medical Association. “We don’t have enough data about them. We don’t understand how they work, how they’re using the information. There’s no liability if these tools get things wrong.”
Pressure for wider access to medical records
Technology companies have also pressed officials to make it easier for health apps to retrieve medical records from the national electronic exchange network known as TEFCA.
At a formal meeting in May, executives argued that an app requesting records should be treated as though it were acting “in the shoes” of the patient. Messages on Slack called for a “frictionless pathway” under which verified users would give consent once and allow an app continuing access to their records.
Ryan Howells, a healthcare management consultant who lobbies for the digital health organisation the CARIN Alliance, has promoted the approach. He has pointed to third-party vetting and an industry code of conduct as safeguards, while also sitting on the board of one of the organisations involved in that vetting.
Kristen Valdes, chief executive of a company connecting apps to the records network, wrote in one discussion: “My opinion is that we need to stop protecting patients from themselves. Full stop.”
Ms Valdes later said her comment reflected frustration with what she called the paternalism of a healthcare system that had historically made it difficult for patients to obtain their records. She acknowledged that there would be “good and bad apps”.
Other participants warned that broad access could create security and privacy risks, particularly if the system expanded to thousands of applications. Jason Kulatunga, who runs a medical records platform, said systems should be designed on the assumption that malicious actors would be present.
Ms Gleason has publicly backed the wider use of apps to transfer records, urging doctors to abandon paper files. At a recent event hosted by a think tank supporting Health Secretary Robert F Kennedy Jr’s agenda, she said patients could now verify their identity and press a button to send their medical records to an app of their choice.
“That was all done just — no regulation, no rulemaking — just by challenging industry to work together,” she said.
Mr Kennedy has separately endorsed the use of AI in healthcare, claiming it is “much better informed than any doctor in the country”. The administration’s approach has been welcomed by technology companies, but has left doctors and legal experts questioning whether the drive to expand AI health apps is moving faster than the safeguards intended to protect patients.
