FDA Launches “Elsa”: Agency-Wide Generative AI Tool to Boost Efficiency

On June 2, 2025, the U.S. Food and Drug Administration (FDA) officially launched Elsa, a generative artificial intelligence (AI) tool developed to enhance operational efficiency across the agency. Built in a high-security GovCloud environment, Elsa supports a wide range of FDA staff from scientific reviewers to field investigators in managing complex, data-intensive tasks.

A Modern Leap in Regulatory Operations

Elsa marks a major advancement in the FDA’s digital modernization strategy. Purpose-built for secure internal use, Elsa does not use confidential industry data for training, ensuring regulatory privacy standards are upheld. Its current functions include:

• Summarizing regulatory documents
• Reviewing clinical protocols
• Analyzing adverse events and safety profiles
• Comparing product labels for compliance
• Generating code for nonclinical database systems

Built Securely, Designed for Scale

Elsa was developed within a FedRAMP-compliant environment and is fully integrated into the FDA’s internal systems. During pilot testing, scientific reviewers validated its capabilities. Following that success, the tool is now being deployed agency-wide.

What’s Next for Elsa?

While Elsa currently focuses on summarization, writing, and coding, the FDA plans to expand its use to include broader data processing capabilities and additional generative AI applications in the future.

Implications for Regulated Industries

Elsa’s rollout is especially relevant for sectors involving clinical evaluations and safety reviews, such as:

• Pharmaceuticals
• Biologics
• Medical devices
• Infant formula
• Other FDA-regulated products

A Responsible Path to AI Adoption

The launch of Elsa underscores the FDA’s commitment to using AI responsibly enhancing efficiency while maintaining strict data security and regulatory compliance.

For more updates on AI-driven innovation in regulatory science, follow RegASK