Lessons from the FDA for AI
Regulators exploring mandatory pre-deployment AI assessment frameworks - a live question for Australian policy - can use this as a structured comparative reference.
Key points
- AI Now Institute draws on FDA pharmaceutical regulation to explore the case for ex ante AI oversight.
- The report maps FDA regulatory functions - premarket approval, post-market surveillance, enforcement - onto AI governance gaps.
- Published in mid-2024, it acknowledges that premarket AI enforcement remains a distant prospect in the US.
Summary
The AI Now Institute's August 2024 report uses the US Food and Drug Administration as an analytical lens for AI regulation, arguing the FDA's transformation of the pharmaceutical sector illustrates the benefits of rigorous ex ante oversight attuned to evolving markets. The report maps FDA functions - including premarket approval, post-market surveillance, and enforcement powers - onto corresponding AI regulatory needs, and identifies four practical challenges: scoping the AI market, granting regulators adequate powers, overcoming legal obstacles, and avoiding industry capture. It is offered as a conceptual resource rather than a prescriptive blueprint, and is candid that mandatory premarket AI enforcement remains politically remote in the US context.
Implications for Australian agencies
- Consider Australian teams working on AI regulatory design - particularly DISR and OAIC - could consult this report's FDA function mapping when assessing gaps in existing Australian AI governance mechanisms.
- Monitor The premarket vs. post-market regulatory debate framed here is likely to resurface in international AI standards discussions; worth tracking for how it influences Five Eyes or OECD regulatory convergence.
Implications are AI-generated. Starting points, not advice.
"Lessons from the FDA for AI" Source: AI Now Institute – Publications Published: 1 August 2024 URL: https://ainowinstitute.org/publications/research/lessons-from-the-fda-for-ai The AI Now Institute's August 2024 report uses the US Food and Drug Administration as an analytical lens for AI regulation, arguing the FDA's transformation of the pharmaceutical sector illustrates the benefits of rigorous ex ante oversight attuned to evolving markets. The report maps FDA functions - including premarket approval, post-market surveillance, and enforcement powers - onto corresponding AI regulatory needs, and identifies four practical challenges: scoping the AI market, granting regulators adequate powers, overcoming legal obstacles, and avoiding industry capture. It is offered as a conceptual resource rather than a prescriptive blueprint, and is candid that mandatory premarket AI enforcement remains politically remote in the US context. Implications for Australian agencies: - [Consider] Australian teams working on AI regulatory design - particularly DISR and OAIC - could consult this report's FDA function mapping when assessing gaps in existing Australian AI governance mechanisms. - [Monitor] The premarket vs. post-market regulatory debate framed here is likely to resurface in international AI standards discussions; worth tracking for how it influences Five Eyes or OECD regulatory convergence. Retrieved from SIMS, 18 May 2026.