Healthcare AI Regulation: 200 State Bills, No Federal Law

States have introduced roughly 200 AI-related bills in 2026. Congress has passed none targeting health AI directly. The gap between state-level activity and federal inaction is defining the compliance reality for every health system deploying healthcare AI and healthcare automation tools this year.

State legislatures are converging on four areas: regulating mental health chatbots, requiring patient disclosure when AI assists in care, preventing AI tools from presenting as clinical providers, and governing how insurers use AI in payment and coverage decisions. California's law, effective January 1, 2026, requires chatbots to disclose their AI status and bans mental health chatbots that lack suicide prevention protocols. Kentucky advanced HB 455 through the House to restrict AI use in therapy services.

At the federal level, the FDA's 510(k) process — the pathway most AI medical devices travel — clears roughly 98 percent of submissions. Penn Medicine faculty have called out the mismatch, noting the framework "was developed decades ago for traditional medical devices" and contains "a lot of holes" when applied to modern AI systems. Their core criticism: the process has no mechanism to monitor whether AI tools continue performing properly after deployment, a gap known as model drift.

Meanwhile, 83 percent of healthcare workers surveyed by Morning Brew said AI needs more regulation. That tracks with a broader tension in the industry: some leaders argue regulatory clarity would accelerate adoption by reducing compliance uncertainty, while others warn that inconsistent state rules will leave multi-state health systems navigating an unworkable patchwork.

---

Sources

• What is the state of healthcare AI regulation? — Healthcare Brew
• Holes in federal AI healthcare regulation should be patched, Penn Med faculty say — The Daily Pennsylvanian