A new era of AI begins at the FDA. The U.S. Food and Drug Administration (FDA) has officially deployed agentic AI capabilities across the agency, allowing all staff to voluntarily access advanced AI tools designed to execute complex, multi-step tasks.

The agency communicated this information on December 1, releasing a press statement where they indicated that this initiative builds on the success of Elsa, the FDA’s large language model (LLM)-powered assistant introduced in May, now reportedly used by more than 70% of staff.

What is agentic AI?

Agentic AI refers to AI systems that can plan, reason, and take action autonomously to achieve defined goals. The agency states that they are built with built-in governance protocols that ensure human oversight and reliable results.

To better understand agentic AI in pharma, we recommend the following articles:

• Agentic AI in Pharma, Part 1: Foundations for automating regulatory content
• Agentic AI in Pharma, Part 2: Memory, learning, and real-world integration

From generative assistants to multi-step AI agents

Unlike traditional AI tools focused on single prompts or tasks, agentic AI enables FDA staff to build workflows that involve multiple AI models working in sequence. These workflows will assist with a range of regulatory operations, including:

• Pre-market review support
• Post-market surveillance
• Compliance inspections
• Meeting coordination and management
• Administrative processing

The FDA has also launched an “Agentic AI Challenge”, encouraging internal teams to create innovative agentic solutions for real-world use cases. Winning entries will be showcased at the FDA Scientific Computing Day in January 2026.

“FDA’s talented reviewers have been creative and proactive in deploying AI capabilities. Agentic AI will give them a powerful tool to streamline their work and help them ensure the safety and efficacy of regulated products,” said Chief AI Officer Jeremy Walsh.

Why is this good news for Narrativa and our clients?

This announcement is a landmark moment for the entire regulatory technology (RegTech) ecosystem and a powerful signal for the future of agentic AI in life sciences. At Narrativa, we have long advocated for AI that goes beyond single outputs to support fully integrated, compliant, and traceable workflows.

Our platform, with solutions like Clinical Atlas, Narrative Pathway, and R-Developer for TLFs, already helps life sciences companies automate regulatory documentation at scale. Now, with the FDA actively adopting and endorsing agentic AI, we anticipate a more streamlined and supportive environment for AI-driven submissions.

Trust, traceability, and compliance are aligned

The FDA’s decision to deploy agentic AI in a high-security GovCloud, where data from regulated industries is not used for training, mirrors Narrativa’s own commitment to privacy and security. This alignment underscores that responsible AI can meet the stringent demands of regulatory bodies while driving massive gains in efficiency, accuracy, and throughput.

The road ahead

For Narrativa and our clients, the FDA’s move validates the direction we have already taken: building AI tools that are composable, secure, and built for regulated environments. It also sets the stage for deeper collaboration with agencies and increased confidence in AI-powered automation.

In short, the FDA’s adoption of agentic AI is not just a technological update. It is an open door to a smarter, faster, and more scalable regulatory future. And Narrativa is ready to lead the way.