FDA Embraces AI: Generative System to Be Fully Operational by June's End

Key points:

1.The United States Food and Drug Administration (FDA) has officially completed its first AI-assisted scientific review pilot.

2.The entire AI system across all departments will be fully deployed by June 30, 2025, covering all centers.

3.The goal of AI applications is to reduce the repetitive work of reviewers and improve the efficiency of evaluating new therapies.

4.The system is being jointly promoted by the FDA's newly appointed Chief AI Officer Jeremy Walsh and Senior Information Officer Sridhar Mantha.

5.Further optimizations will continue to be made to ensure information security and meet the business requirements of each center.

The U.S. Food and Drug Administration (FDA) announced on May 8 that it has successfully completed its first pilot program using generative artificial intelligence (AI) for scientific evaluation, and will fully deploy a unified AI system across all its centers by June 30. The initiative, led by FDA Commissioner Martin A. Makary, aims to systematically improve internal scientific review efficiency and reduce the workload of professionals on repetitive tasks.

"We need to highly value the time of scientists and reduce redundant processes. The widespread application of AI technology can significantly accelerate the speed of reviewing new therapies."

At the same time, he emphasized that the deployment operation this round will be based on the basic principles of "fast, unified, and safe," and will comprehensively establish interfaces between the evaluation data platform and AI tools.

The pilot project uses generative AI technology specifically designed for FDA scientists and experts, with a focus on reducing inefficient, repetitive labor. Jinzhong (Jin) Liu, Deputy Director of the Office of New Drug Review at the FDA Center for Drug Evaluation and Research (CDER), said the tool can complete a task that originally took three days in just a matter of minutes, calling it a "transformative technology.

According to deployment arrangements, all FDA centers have received instructions to immediately initiate internal deployment. June 30th is the timeline for the first round of full system launch, at which time each center will carry out AI-assisted work based on a unified platform deeply integrated with FDA internal data structures. After deployment, FDA will continue to optimize functions, expand specific application scenarios, and ensure that each center can customize their usage paths based on their own business characteristics.

The program is led by Jeremy Walsh, the newly appointed Chief AI Officer at the FDA. Walsh has previously overseen major technology deployments at various federal health and intelligence agencies. He is partnered with Sridhar Mantha, who formerly served as the head of the CDER Business Informatics Office. Together, they will collaborate on advancing overall AI deployment, user training, system security, and ongoing performance assessments throughout the agency.

The FDA announced that more details will be released in June and progress updates will be made public. The agency also promised to continuously iterate AI tools through user feedback mechanisms to ensure they are closely aligned with regulatory tasks and assist in the mission of promoting public health.