FDA has published a significant body of guidance specifically addressing AI and machine learning in medical devices. These documents sit across several categories and are scattered across FDA's website with no single index that connects them to the device-level clearance record they apply to.

FDA AI REGULATION: KEY GUIDANCE TIMELINE 2019 2021 2023 2024 2025 Discussion paper: AI/ML-based SaMD — first framework for AI device regulation Action Plan: AI/ML-Based SaMD — 5 action areas including PCCP framework Draft guidance: PCCP for AI/ML-enabled device software functions FINAL: AI Device Lifecycle Management guidance · FINAL: PCCP guidance Most impactful year for AI device regulation — two major final guidances Draft: AI transparency guidance · Draft: Clinical bias guidance · Action plan update
FDA's AI device regulatory framework has evolved significantly since 2019. The 2024 final guidances are the most consequential for current submissions.

The regulatory effect of different document types

Final guidance represents FDA's current thinking and is used in submission review — deviating without explanation creates review friction. Draft guidance signals where FDA is heading but doesn't yet carry full weight. Action plans and discussion papers outline FDA's thinking without establishing specific reviewer standards.

2024 key guidance documents

Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations (Final, 2024). The most significant AI guidance document of recent years. It formalized the PCCP framework, provided specific transparency requirements for AI device labeling, and established expectations for ongoing performance monitoring. Submissions that don't address this framework face higher probability of additional information requests.

Predetermined Change Control Plans for Machine Learning-Enabled Device Software Functions (Final, 2024). The PCCP-specific guidance that resolved significant ambiguity from the 2023 draft. PCCPs submitted after this guidance are consistently more detailed and better structured than those under the earlier framework.

2025 key guidance documents

Transparency for AI/ML-Enabled Medical Devices (Draft, 2025). Addresses what AI devices must disclose about their algorithm to clinicians and patients. The final version will have material implications for AI device user interfaces and labeling across all cleared device types.

Considerations for the Design, Development, and Testing of AI/ML-Enabled Medical Devices Intended to Address Clinical Bias (Draft, 2025). Proposes specific requirements for subgroup performance analysis. If finalized in its current form, it would significantly increase the validation burden for AI devices, particularly where clinical datasets are historically skewed toward specific demographic groups.

What the guidance trajectory means for 2026 submissions

Labeling transparency requirements are increasing. Subgroup performance documentation is becoming a standard expectation — even before the bias guidance is finalized, reviewers are requesting it with increasing frequency. Post-market commitments are under increased scrutiny. Build all three into your submission strategy from the start.

ⓘ This article is for informational purposes only and does not constitute regulatory or legal advice. AIFDA Intel is an independent platform and is not affiliated with the U.S. Food and Drug Administration. Consult a qualified regulatory affairs professional before making regulatory decisions.