The U.S. Food and Drug Administration (FDA) has taken another bold step in modernizing its digital infrastructure with the deployment of agentic artificial intelligence capabilities across the agency. Announced on December 1, 2025, this initiative represents a significant enhancement of the FDA’s technological toolkit, designed to support staff with advanced, multi-step, AI-driven workflows.
Agentic AI systems are engineered to plan, reason, and execute complex tasks while maintaining strict internal safeguards, including human oversight. These capabilities allow the FDA to broaden its digital transformation strategy while reinforcing a commitment to responsible AI adoption.
Advancing Operational Efficiency and Scientific Rigor
FDA Commissioner Marty Makary, M.D., M.P.H., highlighted the agency’s commitment to innovation, noting that the expansion of AI tools marks one of the most transformative periods in the FDA’s history. “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,” he stated.
The agentic AI deployment builds upon the success of Elsa an LLM-based tool launched in May, which is now actively used by more than 70% of FDA staff. The widespread adoption of Elsa demonstrated the agency’s readiness to deepen its integration of AI into regulatory processes.
Empowering FDA Teams with Multi-Step AI Capabilities
The newly introduced agentic AI tools are designed to handle more sophisticated, multi-layered assignments. Their applications span a broad range of FDA activities, including:
- Meeting and workflow management
- Pre-market product reviews
- Review validation and quality checks
- Post-market surveillance
- Inspection and compliance tasks
- Administrative operations
These capabilities are optional and available to all staff, providing a flexible yet powerful resource for enhancing productivity and improving the agency’s ability to evaluate products effectively and accurately.
Launching the FDA Agentic AI Challenge
To encourage hands-on innovation, the FDA is launching a two-month Agentic AI Challenge, inviting staff to design and test new AI solutions. These innovations will be showcased at the FDA Scientific Computing Day in January 2026, fostering a culture of experimentation and collaboration.
Chief AI Officer Jeremy Walsh emphasized the transformative potential of this technology, noting that agentic AI offers FDA reviewers “a powerful tool to streamline their work and help them ensure the safety and efficacy of regulated products.”
Security, Integrity, and Trust at the Core
The deployment is built within a high-security GovCloud environment that ensures strict data protection. Importantly, the models do not train on staff inputs or on data provided by regulated industries an essential safeguard for maintaining scientific integrity and protecting sensitive information.
This initiative is part of a broader strategy to embed AI thoughtfully across FDA workflows, with the goal of achieving unprecedented operational efficiency while upholding the agency’s mission to protect public health.
New Hope for NUT Carcinoma: FDA Grants Orphan Drug Status to ZEN-3694
Written by Nare Hovhannisyan, MD