CMS Used a Private Slack Room to Shape the Future of AI Health Apps
A year-long campaign inside a government-run Slack workspace has given major technology companies an unusually direct route into federal health policy, while the Trump administration has pushed Americans toward AI-powered apps for managing records, chronic disease, and medical advice. The setup has drawn scrutiny because it blurred the line between public policy development and industry coordination at the very moment federal officials were promising broader access to digital tools through Medicare and other government channels.
The messages and meetings paint a fast-moving effort to normalize AI health apps as part of routine care. Officials at CMS, FDA, and HHS have met with tech companies, investors, and digital health vendors while publicly advancing the administration’s so-called Health Technology Ecosystem, a plan to embed artificial intelligence and related tools into healthcare workflows. In parallel, federal staff helped create a workspace where industry participants could discuss product design, privacy, data access, and reimbursement pathways away from public view.
That arrangement is drawing concern because the same companies seeking access to patients’ records are also being told they may become part of a government-backed pipeline for distribution and payment.
Industry Got a Direct Line to Federal Health Officials
The private Slack workspace launched in August 2025 and quickly grew into a hub for tech leaders, financiers, and federal officials. Amy Gleason, now chief product officer for CMS’ Office of Health Technology and Products, welcomed participants and directed them to the workspace’s code of conduct and channel list. By the following months, the group had expanded to 1,700 members, though only a small number of doctors, hospital representatives, and patient advocates were present.
Among the organizations represented were Microsoft, Anthropic, OpenAI, Apple, and Google. Other participants included digital wellness companies such as Oura Health, AI-related firms such as Palantir, and venture firms including 8VC. The composition matters because it shows who had the closest access while policy around patient-facing AI products was still taking shape: the vendors, investors, and platform companies with the most to gain from a federal endorsement of digital health tools.
One email from CMS senior policy adviser Morgan Taylor framed the project as an opportunity to “support HHS’ goals” and help FDA shape future guidance on conversational AI. In another setting, Jacob Shiff, CMS’ chief AI and technology officer at the Innovation Center, described the agency’s work as a potential “sales engine” for health apps. He told industry participants that CMS would create a directory of organizations and that the companies with workable products would be the winners.
Medicare Is Being Positioned as a Distribution Channel
CMS also launched a Medicare App Library that currently features only two dozen commercial apps, yet is promoted to millions of older adults and people with disabilities. The agency has signaled to industry that the library could serve as a major route to consumers because beneficiaries may trust products listed on a federal website.
That trust is the commercial lever. When a product appears on medicare.gov, it may carry the appearance of government vetting even when the criteria for inclusion are not fully public. In at least one meeting, Taylor said that if Medicare promoted an app, he would treat it as one he would trust and explore. The result is a public-facing directory that can function as a de facto endorsement channel, especially for patients trying to sort through apps claiming to help with obesity, cancer, diabetes, wearable tracking, or generalized health analysis.
Several of the promoted apps are still early-stage products. One example is Slothwise, a health app launched this year that says it can analyze medical records and health data and offers paid features for $9.99 a month. The app has only a handful of reviews in the Apple and Google stores. One Google reviewer wrote that trust in the product came from its Medicare endorsement.
That is the practical consequence of a federal list paired with a weakly disclosed selection process: patients can mistake placement on a government site for a clinical validation that has not actually been demonstrated.
Privacy, Consent, and the Fight Over Medical Records Access
A separate fault line runs through the debate over data access. Tech vendors have lobbied for a model in which authorized apps can retrieve records from TEFCA, the national network that allows providers to electronically exchange medical records, after a third-party vetting process. Some industry leaders argue that once a patient consents, an app should be treated as if it is standing in the patient’s shoes.
On Slack, that argument was pressed hard. Ryan Howells, who works as a healthcare management consultant and lobbies for the CARIN Alliance, pushed for easier access to records through a “frictionless pathway” for verified users. Kristen Valdes, chief executive of b.well, said the system should stop “protecting patients from themselves.” Others in the workspace worried that broader access could create security failures if the system scales to thousands of apps.
Jason Kulatunga, who runs Fasten Health, warned that privacy and security systems must account for malicious actors, not only well-intentioned companies. His concern is straightforward: once medical records can flow easily into consumer apps, the weakest compliance link can become the breach point.
Regulators Face Questions About Oversight and Legality
The central policy question is whether a government-run Slack room, populated heavily by industry participants and used to shape AI health policy, functions like a federal advisory committee without the formal safeguards that such a body would require. Legal experts who reviewed the messages said the structure resembles an advisory panel because members interact, build consensus, and influence agency direction. Formal advisory committees are generally expected to meet in public and include requirements for independence and fair balance.
CMS said the Health Technology Ecosystem is an open, voluntary technical collaboration. Gleason stated that CMS wants to give Medicare beneficiaries greater access to health information and digital tools while promoting an interoperable healthcare system. FDA declined to comment, and CMS declined to answer several questions about the legality of the Slack workspace and the criteria used to place apps in the Medicare App Library.
The policy stakes extend beyond software design. If reimbursement begins flowing to AI chatbots, wearable-tracking services, and data-analysis tools before robust evidence and clear liability standards exist, patients could be nudged into systems that are easier to market than to medically validate. John Whyte of the American Medical Association captured the clinical concern bluntly: many of these tools are not ready for primetime, and there is still too little data on how they work, how they use information, or who is accountable when they get things wrong.
For now, the administration’s digital health push is moving faster than the public debate around safeguards, consent, and oversight. The unanswered issue is whether a government-promoted ecosystem for AI health apps can scale without turning Medicare into a distribution platform for products that are still early, unevenly tested, and thinly regulated.