# What happened Federal healthcare agencies signaled a faster push to integrate AI into clinical care at the Consumer Technology Association's Health AI+ 2026 event in Washington, D.C. CMS leaders detailed a multi-pronged strategy to make AI tools easier to access, evaluate and pay for in Medicare and Medicaid. The FDA described a discussion paper seeking input on how to regulate generative AI–enabled medical devices.
# CMS priorities and ACCESS model CMS Deputy Administrator Stephanie Carlton, who also serves as the agency's chief clinical AI officer, framed clinical AI as a top strategic priority. A central program is ACCESS — Advancing Chronic Care with Effective, Scalable Solutions — a 10-year, value-based chronic-care model that launched this year and emphasizes technology and AI for large patient populations.
ACCESS targets predictable, recurring, outcomes-based payments for technology used to treat conditions such as diabetes, hypertension, chronic kidney disease, obesity, depression and anxiety. Carlton said more tracks and engagement opportunities for ACCESS are expected, though she did not provide detailed timelines.
Carlton described ACCESS as an early test to generate evidence on whether AI tools can improve measurable health outcomes and reduce total cost of care. She said CMS hopes the model will show "that technology can have a massively deflationary impact on healthcare costs" and that it will help test pathways for paying technology companies alongside clinicians and health facilities.
# Four pillars of CMS's AI approach Carlton laid out four primary pillars guiding CMS's AI work:
- Build public trust in AI used in health care.
- Expand healthcare data sharing and interoperability to support AI tools.
- Create clearer pathways for AI market access and regulation.
- Develop reimbursement frameworks for AI-enabled technologies.
Those pillars are intended to shape both how AI tools enter the market and how they are evaluated for clinical use and payment.
# Regulation and reimbursement questions
Carlton emphasized a practical distinction: some AI capabilities will be broadly available in general apps at low cost, while other tools will perform medical functions that meet statutory standards for payment.
# FDA engagement on generative AI devices
The FDA also described exploring a "competency-based" evaluation framework combining extensive benchmarking and testing to assess safety and effectiveness before authorization and to monitor performance post-deployment.
# Clinician debate and broader context
# What to watch next