FDA issues draft guidance on AI/ML in drug development using real-world data

The FDA has published draft guidance addressing the use of artificial intelligence and machine learning tools in drug development contexts that rely on real-world data. The guidance signals the agency’s intent to establish clearer expectations around model transparency, data provenance, and the validation standards it will apply when sponsors submit AI-assisted analyses in support of regulatory decisions. Notably, the FDA is seeking to distinguish between AI tools used for exploratory purposes and those intended to directly inform labelling or approval submissions, with proportionately higher scrutiny anticipated for the latter. The document is open for public comment, giving industry stakeholders a meaningful opportunity to shape the final framework. Regulatory affairs and RWE teams should review the draft now and begin assessing where their current AI workflows may require additional documentation or methodological justification ahead of formal implementation.