FDA Artificial Intelligence/Machine Learning-Based Software as a Medical Device Action Plan
United States (Federal) | FDA AI/ML Action Plan (2021)
Categories
Key Requirements
Commits FDA to a total product lifecycle (TPLC) approach for AI/ML devices. Five action areas: (1) PCCP guidance, (2) AI device labeling and transparency, (3) regulatory science for ML evaluation and bias identification, (4) real-world performance monitoring, and (5) Good Machine Learning Practices (GMLP) development.
Regulated Parties
AI/ML medical device manufacturers; Software as a Medical Device (SaMD) developers
Enforcement
FDA (Center for Devices and Radiological Health; Digital Health Center of Excellence)
Sources
Notes
Strategic framework that spawned subsequent binding guidance including GMLP Guiding Principles (Oct 2021), PCCP Draft Guidance (Apr 2023), and PCCP Final Guidance (Dec 2024). By 2025, FDA had authorized over 1,000 AI-enabled devices.
Companies Affected
Related Wiki Pages
- $100M, $40M, $20M: What 14 Health Systems Got Back from Their Tech Bets
- 25 Healthcare AI Use Cases — Comprehensive Overview with Examples
- 3 Things Healthcare Leaders Should Know About AI-Native EHRs
- Abridge — AI Decision Support Expansion for Clinicians
- AdvaMed Encourages Risk-Based FDA Approach to Generative AI Medical Devices
Want full access to all legislation details?
Upgrade to a Subscriber plan for unlimited law detail views, the state map, and upcoming effective date tracking.
View Plans