The Food and Drug Administration (FDA) has begun rolling out agentic AI tools agency-wide, extending its use of automation technologies beyond the LLM-based assistant it introduced earlier this year. The capabilities allow staff to orchestrate multiple AI models to support operational, scientific, and compliance workflows, with use remaining voluntary.
Elsa, the LLM tool launched in May, is now used by more than 70 percent of staff, according to internal figures cited by the agency. Program teams have since adapted it for a wider range of review and administrative tasks, paving the way for broader agentic adoption.
FDA Commissioner Marty Makary framed the expansion as part of the agency’s modernization push.
“We are diligently expanding our use of AI to put the best possible tools in the hands of our reviewers, scientists and investigators,” Makary said. “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.”
Initial uses for the agentic tools include meeting management, pre- and post-market reviews, inspection support, and compliance activities. The systems operate within a secure GovCloud environment and, according to the agency, do not train on employee inputs or industry submissions. The rollout also introduces a two-month internal Agentic AI Challenge, culminating in demonstrations at Scientific Computing Day in January.