The FDA is scaling Elsa, its GenAI, to streamline clinical protocols, safety summaries, and internal workflows under tight resource pressure.
The FDA is rolling out Elsa, a secure, in-house GenAI assistant, across the agency to support time-consuming clinical and regulatory tasks. Built on a large language model and hosted in GovCloud, Elsa helps staff perform clinical protocol assessments, summarize adverse events, and process internal data more efficiently. Without using proprietary industry data for training.
Elsa was first piloted with scientific reviewers and proved effective in cutting time spent on evaluations. Its agency-wide deployment comes ahead of schedule and under budget. Helping the FDA cope with the impact of 3,500 job losses and proposed budget cuts. These challenges could have slowed drug approvals and inspections. However, GenAI now helps fill the gap by boosting operational efficiency.
The AI tool is already in use for drug label comparisons, safety summaries, and even generating code for nonclinical databases. Elsa enables faster decisions and allows FDA teams to identify high-risk inspection targets more quickly. Designed to improve—not replace—human expertise, Elsa is positioned as a digital co-pilot that enhances reviewer productivity across FDA centers.
This launch marks the first step in a broader AI strategy at the FDA. Future updates will expand Elsa’s functionality based on employee feedback. Planned upgrades may include advanced analytics and new generative tasks, helping the agency adapt to modern demands while maintaining security and compliance. Elsa is a foundational shift in how the FDA manages growing regulatory complexity with leaner resources.