The US FDA Has Launched The Agentic AI System To Enhance Review, Supervision And Compliance Enforcement.

According to the report on the FDA's official website:
The US Food and Drug Administration (FDA) announced today that it will fully deploy the Agentic AI (autonomous AI) capabilities to all agency staff, aiming to simplify and enhance multi-step, complex regulatory workflows - including pre-market review, post-market safety monitoring, inspection and audit, and compliance processes. These advancements may also draw attention from the tobacco and nicotine industries.
This system allows employees to build multi-model AI workflows, which can execute planning, reasoning, and task execution under human supervision. The tool is voluntary and runs in a secure GovCloud environment, and it will not use data submitted by enterprises for training.
In an email reply to StatNews, an FDA spokesperson stated that this tool is still an "exploratory" technology, and the AI agent will not make any regulatory decisions. He emphasized: "All outputs generated by AI will be reviewed and verified by FDA staff before being incorporated into any official regulatory actions to ensure that AI is only used as an auxiliary tool rather than the decision-making entity."
This comprehensive deployment follows the successful application of the FDA's large language model tool Elsa internally. Since its launch in May this year, over 70% of FDA employees have been using Elsa. At the same time, the FDA has also launched a two-month Agentic AI challenge, and selected projects will be showcased in January 2026.
FDA Commissioner Marty Makary stated that this upgrade is an important step in modernizing the regulatory system; Chief AI Officer Jeremy Walsh pointed out that the system will accelerate and strengthen safety assessment and verification processes in all areas under FDA regulation - including tobacco products.