The FDA has expanded its artificial intelligence initiatives by introducing a new agentic AI program that employees may optionally use. In an announcement, the agency stated that the tool is intended to assist with meeting management, premarket reviews, post-market surveillance, and inspections.
Agentic AI systems operate by allowing different components of a model to perform separate stages of a task, with their outputs integrated through a decision-making process that includes human oversight. The FDA said the new model has not been trained on data submitted to the agency and is hosted in a high-security GovCloud environment.
FDA Commissioner Marty Makary, M.D., said in the announcement, “We are diligently expanding our use of AI to put the best possible tools in the hands of our reviewers, scientists, and investigators.” He also noted that the agency is working to modernize its operations with these systems.
Makary previously stated in May that he intended for the FDA to fully integrate generative AI into its work by the end of June, a plan that preceded the early launch of Elsa, a large language model designed to accelerate review activities. Elsa was criticized after reports that she provided inaccurate responses when asked to identify FDA leaders or summarize information about FDA-approved products. The FDA later said that 70% of its staff had voluntarily used the system.
Earlier this year, large-scale layoffs significantly reduced the FDA’s workforce, including employees on review teams. During this period, the agency missed several target decision dates for potential drug approvals, most recently extending the PDUFA date for Denali Therapeutics’ Hunter syndrome therapy by three months.
On Nov. 21, hundreds of biopharma leaders sent a letter to Makary expressing concerns about the FDA’s ability to function, citing issues including leadership turnover, missed deadlines, extensive layoffs, and what the letter described as the regulator’s “failure to stand up for established vaccine science and important mRNA research.”
In addition to its AI efforts, the FDA recently said it was beginning a hiring spree aimed at adding more reviewers, although no related job listings have appeared on the U.S. government’s official job board. The agency has also introduced a new “national priority” voucher program offering one- to two-month reviews, which has drawn scrutiny from lawmakers and was publicly criticized by George Tidmarsh, M.D., Ph.D., the former head of the FDA’s Center for Drug Evaluation and Research.
The FDA said that agentic AI is “entirely optional” for employees. It will support functions such as managing meetings, conducting pre-market reviews, validating reports, completing inspections, and other administrative tasks. To encourage its use, the FDA will hold a two-month Agentic AI Challenge, with employees presenting their agent-based solutions in January. Both Elsa and the new agentic AI model operate within a secure cloud environment and do not have access to data from regulated companies.
FDA Introduces Agentic AI Program to Support Internal Review and Oversight Activities
The U.S. Food and Drug Administration has announced a major digital transformation initiative with the launch of its Agentic AI program. This forward-thinking technology framework is designed to enhance how the agency reviews submissions, manages internal data, and oversees regulated products. As the FDA expands its modernization strategy, Agentic AI is emerging as a critical tool for improving efficiency and supporting evidence-driven decision-making across the organization.
The newly introduced program will allow the agency’s review teams to automate repetitive tasks, analyze large datasets with greater accuracy, and strengthen oversight processes that protect public health. With Agentic AI integrated into workflow systems, reviewers can access more timely insights, identify risks earlier, and reduce administrative burdens that traditionally slow regulatory progress.
This initiative demonstrates the FDA’s commitment to responsible innovation. By implementing Agentic AI, the agency is not only embracing advanced computational methods but also ensuring its oversight capabilities remain strong in an increasingly complex healthcare and technology landscape.
Experts note that this shift reflects a broader effort across federal agencies to use intelligent automation for quality improvement, data integrity, and transparent regulatory monitoring. As the FDA moves forward, the role of Agentic AI will continue to expand—bringing sharper analytics, improved consistency, and stronger internal infrastructure for everyone.
