The FDA announced it launched a new generative AI platform, Elsa, to support its workforce with an agency-wide tool.
The agency said Elsa is designed to increase operational efficiency across scientific and regulatory functions. Elsa is built within a high-security GovCloud environment. It is available to FDA employees, including scientific reviewers, inspectors, and policy staff, to accelerate tasks.
The FDA said Elsa can be used for clinical protocol reviews, safety assessments, and inspection targeting. The tool is not trained on data submitted by the regulated industry, and all information accessed or generated remains internal to the agency.
The launch, announced Monday, follows an internal pilot with scientific reviewers and comes nearly a month ahead of the FDA’s targeted deployment date of June 30.
“Today marks the dawn of the AI era at the FDA with the release of Elsa, AI is no longer a distant promise but a dynamic force enhancing and optimizing the performance and potential of every employee,” FDA Chief AI Officer Jeremy Walsh said in a news release. “As we learn how employees are using the tool, our development team will be able to add capabilities and grow with the needs of employees and the agency.”
Elsa is powered by large language models and is designed to assist employees with document reading, summarization, label comparisons, code generation, and other language-heavy tasks. According to the FDA, it can already support adverse event reviews and the development of nonclinical databases. The system is expected to evolve as agency staff identify new applications.
The agency framed the launch as a foundational step in its broader AI strategy. Future updates to Elsa may include expanded data processing capabilities and new generative AI functions to support additional operational areas.
