A fresh legal roundup outlines where AI (artificial intelligence) policy stands now—from FDA (US Food and Drug Administration) frameworks to an active wave of state laws.

Summary

A new National Law Review overview surveys the fast-moving U.S. healthcare AI landscape, pulling together recent federal actions and a surge of state legislation. Highlights include FDA guidance activity for AI in drug development, a new RFC (request for comments) on how to measure real-world performance of AI-enabled medical devices, and broad state action on AI bills in 2025. Read the brief. (National Law Review)

What happened

• Federal guidance, drugs: FDA published draft guidance on using AI to support regulatory decision-making for drugs and biologics, offering a credibility framework for AI models and their “context of use.” See FDA page. (U.S. Food and Drug Administration)
• Federal inquiry, devices: FDA opened an RFC on how to measure and evaluate real-world performance of AI-enabled medical devices; comments are due December 1, 2025. Details here. (FDA Law Blog)
• Whole-of-government signal: The White House released America’s AI Action Plan, a policy roadmap with cross-agency actions that includes healthcare implications. Read the plan. (The White House)
• State surge: In 2025, all 50 states, D.C., Puerto Rico, and the U.S. Virgin Islands introduced AI bills; 38 states have adopted or enacted ~100 measures so far, some touching health and insurer use of AI. NCSL (National Conference of State Legislatures) tracker. (NCSL)

Key numbers and dates

• Jan 6, 2025: FDA issued draft guidance on AI in drug/biologic development. Guidance. (U.S. Food and Drug Administration)
• Jul 10, 2025: White House released America’s AI Action Plan. PDF. (The White House)
• Sep 30, 2025: FDA posted RFC on measuring AI-enabled device performance; comments due Dec 1, 2025. Docket summary. (FDA Law Blog)
• 2025 state action: 38 states enacted ~100 AI measures; all 50 states introduced bills. NCSL tracker. (NCSL)

What to watch next

• Whether FDA’s drug and device AI efforts converge into coordinated expectations for model validation, transparency, and real-world monitoring. FDA overview pages. (U.S. Food and Drug Administration)
• State AI bills that specifically address healthcare, payers, and provider uses of AI (e.g., disclosures, bias testing, or record-keeping) as 2025 measures are implemented. NCSL tracker. (NCSL)
• Follow-on FDA guidance or pilots stemming from the device RFC on measuring performance in real-world use. RFC note. (FDA Law Blog)

Why this matters for you

Teams building or buying AI for clinical development, safety, or market access need clear documentation for model purpose, data lineage, validation, monitoring, and human oversight. Align your governance playbook to FDA’s credibility principles, and be ready to show how you measure real-world performance—not just accuracy in a sandbox—especially for any tools touching evidence used in submissions.

Legal, digital, and medical leads should track state-level AI laws alongside federal guidance. Requirements on disclosure, bias testing, and audit trails can vary by state, so codify standard operating procedures that cover consent, explainability, and record-keeping. The organizations that can demonstrate trustworthy AI—from protocol to patient support—will speed internal approvals and reduce friction with regulators, payers, and provider partners.

Source roundup: National Law Review overview on U.S. healthcare AI policy; FDA guidance and RFC notices; White House America’s AI Action Plan; and NCSL state legislation tracker. (National Law Review)