FDA Introduces Agentic AI Platform for Staff Workflows

Dec 3, 2025
The U.S. Food and Drug Administration has launched an agentic AI system for internal use, allowing staff to automate complex tasks such as reviews, inspections, and meeting management. The system is hosted on a secure GovCloud environment and will operate under human oversight.

The U.S. Food and Drug Administration (FDA) has introduced an agentic artificial intelligence platform designed to assist employees with complex, multistep tasks. Staff can use the system for meeting management, pre-market and post-market reviews, surveillance, inspections, and administrative functions.

The agency described the rollout as exploratory and confirmed that the system will not make regulatory decisions or replace human judgment. The AI operates within a high-security GovCloud environment and does not train on data submitted by regulated industries, ensuring confidentiality of sensitive information.

Agentic AI models are built to pursue defined goals through coordinated actions between components, with human oversight guiding the final outputs. The FDA’s staff have been invited to test and build solutions using the system over the next two months, with demonstrations planned for a Scientific Computing Day in January.

This initiative follows the earlier release of the FDA’s generative AI model, Elsa, which has been voluntarily adopted by more than 70% of the agency’s workforce. The new agentic platform expands the agency’s use of artificial intelligence as it seeks to improve efficiency in regulatory and scientific workflows.

We hope you enjoyed this article

Free newsletter

Enterprise AI Brief

Weekly report on AI business applications, enterprise software releases, automation tools, and industry implementations.

Whitepaper

Tensordyne Napier: What If One Rack Could Do the Work of Nine?

Tensordyne

This Tensordyne whitepaper presents Napier, an inference-focused AI processor and rack-scale system based on the company’s TDN Math logarithmic number system. It examines infrastructure requirements for large mixture-of-experts and agentic models, compares major inference architecture approaches, and details the TDN AIP processor, TDN72 pod, TDN Link fabric, and Napier Ultra configuration. The paper reports simulation-based performance, cost, and accuracy-validation results, including Tensordyne’s projected comparison of one Napier rack with a nine-rack Nvidia Rubin plus Groq deployment; the chip is reported as taped out and in fabrication.

Read more
Free, six days a week

Daily AI Brief: the AI news that matters, in your inbox.