FDA launches generative AI tool Elsa to streamline operations and accelerate scientific review
The FDA has launched Elsa, an AI tool that speeds up document search and review while protecting sensitive data. Elsa supports faster clinical reviews and inspection targeting.
FDA Launches Elsa: A New AI Tool to Streamline Agency Operations
The U.S. Food and Drug Administration (FDA) has introduced Elsa, a generative AI tool aimed at improving efficiency for agency employees, including scientific reviewers and field investigators. This marks the FDA’s first significant adoption of generative AI technology, signaling a shift toward smarter and faster workflows.
Built within a secure GovCloud environment, Elsa enables staff to quickly search and summarize internal documents without risking exposure of sensitive research or proprietary data. This ensures that confidential information remains protected while improving access to critical insights.
Transforming Workflows with AI
Elsa goes beyond being a simple chatbot. It helps reduce the time spent on repetitive and routine tasks, often a bottleneck in review processes. Importantly, the AI models do not learn from data submitted by regulated industries, which safeguards confidential research.
Jeremy Walsh, the FDA’s Chief AI Officer, noted that Elsa marks the beginning of the agency’s AI era. The tool is expected to evolve based on employee feedback, expanding its capabilities to better serve the agency’s diverse needs.
Accelerating Reviews and Inspections
The FDA is already leveraging Elsa to speed up clinical protocol reviews, shorten scientific evaluation times, and pinpoint high-priority inspection targets. FDA Commissioner Dr. Marty Makary highlighted that the rollout was completed ahead of schedule and under budget, thanks to strong collaboration across internal teams.
Following a successful pilot with scientific reviewers, the agency aimed to deploy Elsa across all centers by June 30, a goal achieved earlier than planned.
How Elsa Supports Operational Efficiency
- Summarizes adverse events to aid safety profile assessments
- Performs quicker label comparisons
- Generates code to assist in database development for nonclinical applications
These features represent the first phase of a broader plan to embed AI into FDA workflows, boosting productivity while maintaining strict data security and compliance with agency policies.
Future Enhancements and Agency Collaboration
The FDA plans to expand Elsa’s functionality by improving usability, enhancing data processing, and refining generative AI functions. Outputs will be customized to meet the specific requirements of different FDA centers without compromising security.
This agency-wide rollout was led by Jeremy Walsh and Sridhar Mantha, who bring experience from federal health and intelligence sectors, as well as business informatics within the FDA’s Center for Drug Evaluation and Research (CDER).
Operations professionals interested in adopting AI tools to improve workflows can learn more about practical AI applications and training at Complete AI Training.
```
