TLDR: The U.S. Food and Drug Administration (FDA) has launched ‘Elsa,’ an internal generative AI tool designed to significantly enhance the efficiency and speed of its food safety and recall processes. Rolled out in early June 2025, Elsa assists FDA staff by summarizing reports, identifying high-risk inspection targets, and accelerating data analysis, aiming to reduce the time it takes for critical safety information to reach the public.
The U.S. Food and Drug Administration (FDA) has ushered in a new era of food safety oversight with the introduction of ‘Elsa,’ a sophisticated internal artificial intelligence (AI) tool. Launched in early June 2025, with an aggressive timeline for agency-wide integration by June 30, Elsa is poised to revolutionize how the FDA identifies and responds to food safety issues, including the critical process of food recalls.
Elsa, a generative AI tool powered by large language models (LLMs), is designed to assist FDA staff across various functions, from scientific reviewers to investigators and compliance officers. Its primary purpose is to streamline document-heavy tasks and improve the speed and efficiency of internal workflows. FDA Commissioner Marty Makary, M.D., M.P.H., emphasized the tool’s potential, stating, ‘I was blown away by the success of our first AI-assisted scientific review pilot. We need to value our scientists’ time and reduce the amount of non-productive busywork that has historically consumed much of the review process. The agency-wide deployment of these capabilities holds tremendous promise in accelerating the review time for new therapies.’
Key applications of Elsa include summarizing vast volumes of adverse event reports to quickly identify critical safety signals, assisting in the review of complex clinical protocols, and identifying higher-risk inspection targets based on historical data. The tool can also generate structured database code for internal use and perform faster label comparisons. These capabilities are expected to significantly shorten the timeline for serious food recalls, which can currently take three to five weeks or longer to be officially classified and reach the public. Jinzhong (Jin) Liu, Deputy Director, Office of Drug Evaluation Sciences, Office of New Drugs in FDA’s Center for Drug Evaluation and Research (CDER), lauded Elsa as a ‘game-changer technology that has enabled me to perform scientific review tasks in minutes that used to take three days.’
While Elsa is a powerful assistant, it is crucial to note that it is an internal tool that supports staff and does not make final regulatory decisions or engage directly with consumers. It is built within a high-security government cloud (GovCloud) environment, ensuring that sensitive FDA data remains protected and that the models do not train on proprietary industry submissions. Jeremy Walsh, FDA’s Chief AI Officer, highlighted this shift, stating, ‘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.’
However, the deployment of Elsa is not without its considerations. Some FDA staff have reportedly flagged concerns regarding the tool’s accuracy with large datasets and the ongoing need for robust human oversight. Reports have also indicated that Elsa may have limitations, such as character limits for review and the potential for ‘hallucinations,’ where it could produce false information. The agency acknowledges that Elsa is still in its early stages of rollout and plans to continue assessing performance, gathering user feedback, and refining features.
Also Read:
- NIST Releases Key Guidelines for Integrating Artificial Intelligence into Fire Service Electronic Safety Equipment
- PepsiCo Integrates Salesforce Agentforce AI for Enhanced Operations and Market Reach
Elsa’s introduction is a significant step in the FDA’s broader digital transformation strategy, reflecting a commitment to adopting responsible AI tools to enhance its public health mission. As AI becomes increasingly prevalent in the food system, from formulation to supply chain modeling, the FDA’s proactive integration of such technology aims to ensure faster, more consistent responses to emerging safety issues and regulatory needs.
