The U.S. Food and Drug Administration has rolled out agentic AI tools to every agency employee, aiming to simplify complex, multi-step regulatory work — from pre-market reviews and post-market surveillance to inspections and compliance tasks that matter to the tobacco and nicotine industries.

These tools let staff assemble multi-model AI workflows that can plan, reason, and carry out tasks while remaining under human supervision. Use of the systems is optional; they run in a secure GovCloud environment and are not trained on data submitted by industry.

The move builds on the success of Elsa, an internal large language model introduced in May that is now used by more than 70% of FDA staff. The agency is also launching a two-month Agentic AI Challenge, with selected projects set to be showcased in January 2026.

FDA Commissioner Marty Makary called the upgrades a major step in modernizing regulatory operations. Chief AI Officer Jeremy Walsh noted the potential to speed up and strengthen safety assessments across all FDA-regulated sectors, including tobacco.

This article was adapted from an original report published on tobaccoreporter.com. All rights belong to the original publisher.