FDA taps generative AI to tackle scientific data deluge and boost efficiency

The FDA‘s landmark decision to deploy generative AI across its entire organization by June 2025 represents a major shift in how U.S. regulatory agencies are embracing artificial intelligence. This accelerated adoption follows a successful pilot program in scientific review processes and signals how federal agencies are reimagining workflows to address the growing complexity of regulatory oversight and exponential growth in scientific data that requires expert analysis.

The big picture: The FDA plans comprehensive enterprise-wide AI implementation in just over a year, positioning the agency at the forefront of federal digital transformation efforts.

• The agency has established a new Center of Excellence and AI Governance Board to oversee the rollout and has already deployed secure AI platforms for internal testing.
• According to the FDA’s announcement, the technology will assist with document review, regulatory compliance tracking, and scientific analysis.

What they’re saying: FDA officials report dramatic efficiency improvements from their initial AI pilot programs.

• “This is a game-changer technology that has enabled me to perform scientific review tasks in minutes that used to take three days,” said Jinzhong Liu, Deputy Director in FDA’s Center for Drug Evaluation and Research.
• FDA Commissioner Martin Makary emphasized the need to “reduce the amount of non-productive busywork that has historically consumed much of the review process.”

How it works: The first successful pilot used generative AI to review documents related to Investigational New Drug applications, which the agency now plans to scale across all centers.

• Future efforts will focus on improving usability, broader document integration, and customizing outputs to meet the specific needs of different FDA centers.
• The agency promises strong information security protocols and continuous performance evaluation to refine the system based on staff feedback.

Key safeguards: The FDA emphasized that its AI implementation will maintain human oversight and decision-making authority.

• The agency publicly committed that AI models will not make independent decisions, with all outputs continuing to undergo human review.
• This human-in-the-loop approach aims to preserve safety, privacy, and decision integrity throughout the AI adoption process.

What’s next: The FDA will release more information and updates on this initiative in June when the enterprise-wide deployment begins.