The FDA's commitment to artificial intelligence (AI) faces uncertainty following key leadership changes, particularly with the departure of Dr. Marty Makary, a strong advocate for a centralized AI strategy. Under his guidance, the FDA initiated the implementation of a large language model named Elsa, aimed at enhancing efficiency in reviews and evaluations by automating support tasks.
Despite a significant increase in reported AI use cases—up 148% between 2024 and 2025—the future direction of AI at the FDA is now in question. Acting Commissioner Kyle Diamantas maintains that AI remains a priority, but the exit of top officials has raised concerns about the governance structure for AI initiatives. Experts suggest this could lead to a fragmented approach similar to what existed prior to Makary's tenure, potentially hindering efforts toward transparency and cohesive policy-making regarding AI.
Internally, the FDA's use of AI, particularly through the Elsa tool, has been evolving. Designed to assist staff without making final decisions, it helps streamline tasks by providing access to relevant information tailored for specific roles. However, there is still a notable lack of transparency about how AI is integrated into regulatory processes, which could impede collaboration with the industry.
Looking ahead, experts emphasize the need for clearer communication from the FDA regarding its AI applications and a more agile approach to policy development. While traditional regulatory processes can ensure stability, they may also slow down necessary adaptations in a rapidly evolving technological landscape. Balancing structure with flexibility will be crucial as the FDA navigates its AI strategy moving forward.