From Pilot to Policy: The FDA’s Rapid March Toward AI-Powered Drug Reviews
When FDA reviewers discovered that a generative-AI pilot could slash a 3-day document review task to mere minutes—a time-savings of 99%!—it became clear that artificial intelligence is no longer an aspirational tool but a present-day catalyst for regulatory transformation. That single statistic captures why FDA is fast-tracking AI: the volume and complexity of regulatory submissions continue to rise, while patient need and sponsor expectations demand ever-faster answers. This post unpacks how the US FDA is incorporating AI into its scientific review process, what it means for Sponsors, and what to watch out for.
An Ambitious Deadline . . . and Why It Matters
Commissioner Martin A. Makary1 has ordered every FDA product center to deploy AI-based scientific-review programs now, with full integration on a secure, unified platform by June 30, 2025.2 The mandate turns years of exploratory workshops into an agency-wide operational requirement. For sponsors, that timeline signals a near-term shift in how new drug applications (NDAs) and biologic licensing agreements (BLAs) will be read, triaged, and discussed by the FDA review team.
What the First Pilot Proved
The Center for Drug Evaluation and Research’s inaugural pilot2 used generative AI to parse investigational new-drug applications (INDs). Reviewers reported that with the use of AI repetitive extraction of data, cross-referencing, and summarization tasks “now take minutes instead of three days,” freeing scarce scientific talent for higher-order analysis. Dr. Makary’s takeaway: cut the “non-productive busywork” so experts can focus on true benefit–risk questions—an efficiency gain with an enormous public health upside. The success of that pilot is the impetus for accelerating the broader adoption of AI across the agency. The policy goals that have come out of this pilot program include
- Operationalizing generative AI across all centers on one secure platform
- Compressing review-cycle busywork by orders of magnitude, freeing FDA reviewers for judgment-driven tasks
- Embedding human oversight, robust validation, and transparency to safeguard public trust
Guardrails First, Algorithms Second
Despite the speed of implementation, the FDA is adamant that humans remain in the loop. AI outputs are advisory, and decision authority stays with credentialed reviewers. The rollout is being overseen by newly appointed Chief AI Officer Jeremy Walsh and digital strategist Sridhar Mantha. Behind the scenes, the FDA has already logged multiple AI use cases in HHS’s 2024 inventory,3 giving it a knowledge base for scaling responsibly.
Where AI Will Add Immediate Value
Early use cases extend well beyond faster reading comprehension and show the utility of AI to improve efficiency at multiple steps in the review process:
- Protocol & results summarization: condensing hundreds of pages of text and data tables into reviewer-ready abstracts
- Safety-signal detection: flagging safety signals across spontaneous reports, trials, and literature
- Statistical robustness checks: stress-testing endpoints and subgroup claims
- Benefit–risk synthesis: pulling structured and unstructured data into a single vantage point
For sponsors, this means that meticulous consistency across modules will be rewarded, and data anomalies will surface earlier in the review process, potentially reducing late-cycle information requests.
A New Kind of Regulator-Industry Handshake
“Regulators aren’t just observers—they’re becoming digital collaborators,” notes former Pfizer executive Fouad Akkad. As FDA rolls out this AI initiative, Sponsor teams must mirror that investment in AI. For example, if the Sponsor knows that the Agency is using AI-review, they should consider writing their submissions for easier AI-assisted consumption—think structured metadata, standardized terminology, and machine-readable tables. In this way, Sponsors and the FDA become collaborators, and working together can produce benefits for both. The bottom line is that companies that cultivate AI fluency across their submission materials will likely hold a competitive edge.
Both Sponsors and the FDA Need to Manage Risk
The FDA’s own guidance underscores that faster computing does not equal automatic objectivity. Both Sponsors and the FDA need to consider the common risks related to AI use, such as bias in the algorithm(s), awareness of what knowledge base the AI is using and how it was built, and privacy vulnerabilities.
Conclusion
Implementation of AI-review at FDA has the potential to greatly improve the efficiency of application reviews. For sponsors, this evolution demands proactive adaptation—both in how data are formatted and packaged, as well as how they are communicated. The upside is clear: patients may gain faster access to safe, effective therapies, and sponsors gain predictability in an increasingly complex evidence landscape. How is your organization preparing for an AI-enabled dialogue with regulators?
Aaron Csicseri, PharmD
Vice President, Scientific Services
Dr. Csicseri joined the ProEd Regulatory team in November 2017 as a scientific director, responsible for scientific leadership, content development, strategic input, and effective moderation of team meetings. Aaron has extensive experience guiding Sponsor teams through the AdCom preparation process. He received his PharmD at the University of Buffalo, where he studied the clinical curriculum. Aaron has 10+ years of experience as a medical director/clinical strategist in the accredited medical education field (CME), as well as in the non-accredited PromoEd sphere. Over the past 8 years, he has been supporting Sponsors in their preparations for FDA and EMA regulatory meetings in a wide variety of therapeutic areas. Aaron is based in Grand Island, NY, just outside Buffalo.
References
- American Hospital Association. FDA commissioner talks of shortening drug approval process, benefits of AI [press release]. May 6, 2025. https://www.aha.org/news/headline/2025-05-06-fda-commissioner-talks-shortening-drug-approval-process-benefits-ai
- US Food and Drug Administration. FDA announces completion of first AI-assisted scientific review pilot and aggressive agency-wide AI rollout timeline [press release]. May 8, 20205. https://www.fda.gov/news-events/press-announcements/fda-announces-completion-first-ai-assisted-scientific-review-pilot-and-aggressive-agency-wide-ai.
- Assistant Secretary for Technology Policy. HHS AI Use Case Inventory 2024. https://www.healthit.gov/hhs-ai-usecases?search_api_fulltext=FDA.
- US Food and Drug Administration. Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products [Guidance Document]. January 2025. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-use-artificial-intelligence-support-regulatory-decision-making-drug-and-biological.