New FDA Guidance on the Use of AI to Support Regulatory Decisions
The US Food and Drug Administration (FDA) has just taken a historic step by releasing its first-ever draft guidance on using artificial intelligence (AI) for regulatory decision-making related to the safety, effectiveness, or quality of drugs and biologics.1 Published in January 2025, this framework comes at a pivotal time, as AI’s potential to transform healthcare continues to accelerate.
Since 2016, the FDA has reviewed more than 500 drug and biological product submissions containing AI functions, tools, or algorithms—a clear indicator of how integral this technology has become in drug development.
Instead of a one-size-fits-all approach, the FDA draft guidance proposes a practical, risk-based model that considers the “context of use”—how and where an AI model is applied. The potential consequences of AI use within a drug or biologic submission can vary widely, and the new framework acknowledges that high-risk applications require more stringent validation.
Guidance Examples of AI Use Cases Intended to Support Regulatory Decision-Making:
- Reducing Animal-Based Studies: AI can reduce the number of pharmacokinetics (PK), pharmacodynamics, and toxicologic studies involving animals.
- Predictive Modeling for Clinical Pharmacokinetics: AI can facilitate predictive modeling for clinical PK and/or exposure-response analyses.
- Data Integration from Various Sources: AI can integrate data from sources, including natural history studies, clinical trials, genetic databases, clinical studies, social media, and registries, to enhance understanding of disease presentations, heterogeneity, predictors of progression, and recognition of disease subtypes.
- Processing Large Data Sets: AI can process and analyze extensive datasets, including real-world data or data from digital health technologies, for the development of clinical trial endpoints or the assessment of clinical outcomes
- Postmarketing Adverse Event Reporting: AI can be used to identify, evaluate, and process adverse drug experience information for postmarketing surveillance.
- Facilitating Manufacturing Conditions: AI can optimize the selection of manufacturing conditions in pharmaceutical production processes.
Critical Aspects of AI Regulation
The FDA’s draft guidance highlights several critical areas sponsors need to address when incorporating AI into regulatory submissions, each emphasizing the agency’s focus on safety, efficacy, and equity in AI-driven innovation:
- Dynamic vs “Locked” AI Models: The agency differentiates between “locked” or static AI (which doesn’t change after deployment) and “dynamic” models (which update in real time). This distinction matters for regulatory review and postmarketing surveillance because continuous learning may alter performance beyond the initial approval.
- Real-World Evidence (RWE): The FDA’s increasing acceptance of RWE ties closely to AI’s ability to process large-scale, real-world datasets (eg, electronic health records). The new guidance could open doors for more AI-driven RWE studies supporting drug approvals, complementing efforts supported by the 21st Century Cures Act.2
- Ethical & Bias Concerns: Ensuring diverse, representative, training data is paramount. The FDA is keen on preventing AI-driven decisions that inadvertently disadvantage certain populations. Sponsors will need to demonstrate how they’ve mitigated bias to protect patient safety and equitable access to new therapies.
- Early Engagement: The FDA recommends early engagement (eg, pre-IND meetings, Type C meetings) for sponsors planning to use AI. Early discussions can clarify expectations and identify potential pitfalls, preventing costly delays down the road.
What This Means for the Industry
The FDA’s framework is the result of extensive stakeholder feedback:
- 800+ comments on prior discussion papers
- An FDA-sponsored expert workshop at Duke Margolis Institute
- Ongoing input from manufacturers, tech developers, and academia
Still, debate persists on balancing safety with innovation. How strictly should AI be regulated before it stifles creative solutions? The new guidance strives for a middle path—one that fosters innovation while maintaining patient protections.
What’s Next?
The FDA is seeking public comment on the draft guidance over the next 90 days.3 They’re particularly interested in feedback on how well the guidance aligns with industry experience and whether the engagement options for sponsors are sufficient. For companies looking to incorporate AI into their drug development process, the FDA strongly encourages early engagement to set appropriate expectations and identify potential challenges upfront.
As we continue to see AI reshape healthcare innovation, these guidelines represent a crucial step toward ensuring that advancement aligns with FDA expectations and doesn’t come at the cost of patient safety. The FDA’s framework provides a clear path forward for responsible AI implementation to support regulatory decisions and application of new technologies, setting the stage for the next generation of medical innovations.
What are your thoughts on the FDA’s approach to AI regulation in healthcare? Share your perspectives in the comments below.
Aaron Csicseri, PharmD
Vice President, Scientific Services
Dr. Csicseri joined the ProEd 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:
- US FDA. Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products; Draft Guidance for Industry and Other Interested Parties. January 2025. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-use-artificial-intelligence-support-regulatory-decision-making-drug-and-biological
- US FDA. 21st Century Cures Act. January 2020. https://www.fda.gov/regulatory-information/selected-amendments-fdc-act/21st-century-cures-act
- Regulations.gov. Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products; Draft Guidance for Industry; Availability; Agency Information Collection Activities; Proposed Collection; Comment Request. January 2025. https://www.regulations.gov/docket/FDA-2024-D-4689/document