FDA Deploys New AI Tool Under Efficiency Scheme

Share

FDA says it has deployed advanced “agentic AI” capabilities to all agency employees, marking a major expansion of its artificial intelligence (AI) tools for regulatory and scientific work. According to the agency, Agentic AI systems are designed to plan, reason, and execute multi-step tasks autonomously while incorporating human oversight to ensure reliable results. The tools, which staff use voluntarily, can assist with complex workflows including pre-market reviews, post-market surveillance, inspections, compliance tasks, administrative functions, and meeting management, it says.

“Expanding our use of AI puts the best possible tools in the hands of our reviewers, scientists and investigators,” FDA commissioner Marty Makary is quoted in a release as saying. “There has never been a better moment in agency history to modernize with tools that can radically improve our ability to accelerate more cures and meaningful treatments.”

The rollout follows an earlier deployment in May of Elsa, a large language model-based AI tool, which has been voluntarily used by over 70% of FDA staff, Makary says. Agency teams have refined Elsa since its introduction to better integrate it into day-to-day operations.

As part of the new rollout, the FDA is launching a two-month “Agentic AI Challenge” in which employees can develop AI solutions and showcase them at the FDA Scientific Computing Day in January 2026.

The AI models operate entirely within a high-security GovCloud environment, ensuring that sensitive research and data handled by the FDA are protected, Makary says, adding that the models do not train on input data or any data submitted by regulated industries.

The new deployment represents a key step in FDA’s ongoing plan to embed AI across workflows and enhance operational efficiency across the agency.

Read more