Shift in FDA Leadership and AI Strategy

Following the resignation of Marty Makary, former director who spearheaded AI integration at the U.S. Food and Drug Administration (FDA), significant changes in regulatory policy direction are being observed. Furthermore, the successive departures of key operational leaders, including Jeremy Walsh, Chief AI Officer (CAIO), and Sridhar Mantha, interim Chief Information Officer (CIO), have created ambiguity in the agency's governance. While Kyle Diamantas, acting director, emphasizes that AI implementation remains a top priority, the absence of a central command structure inevitably weakens the FDA's ability to execute its comprehensive AI roadmap. This poses a serious risk of causing significant inconsistencies in regulations in the short term.

Slowdown in the Implementation of In-House AI System 'Elsa'

The implementation of 'Elsa,' the FDA's in-house large language model (LLM)-based administrative support tool developed to enhance internal operational efficiency, is also likely to be delayed. Elsa originated from the Center for Drug Evaluation and Research's (CDER) Retrieval-Augmented Generation (RAG) system, 'CDER GPT,' and was meticulously designed to control the hallucination phenomenon associated with AI. It was expected to significantly assist reviewers in summarizing extensive clinical reports and shortening review periods. However, with the loss of momentum, there is a growing possibility that the agency will revert to its previous fragmented environment, where individual departments use separate tools. This could exacerbate administrative bottlenecks within the FDA, ultimately leading to longer drug review times.

Return to Traditional Regulatory Procedures and Industry Uncertainty

Under the leadership of Acting Director Diamantas, the FDA is seeking to 'normalize' regulations through the formal development of guidelines rather than through journal publications or informal press releases. This is a response to the perception that Makary's intuitive and flexible policy communication style, which did not follow formal administrative procedures, caused confusion. However, developing and validating formal guidelines takes at least a year, which could create a bottleneck that prevents the FDA from keeping pace with the rapidly evolving AI market. Pharmaceutical companies and contract research organizations (CROs) are likely to face difficulties in developing business plans as they must validate AI models in the absence of official standards.

Medium- to Long-Term Impact on the AI-Driven Drug Development Market

While this leadership change is not expected to directly regulate the use of AI technology by pharmaceutical companies, the delay in developing clinical trial validation guidelines could have an indirect impact. The global AI in Drug Discovery market is projected to grow rapidly to approximately $2.9 billion in 2026, potentially reaching up to $5.09 billion. Bio-tech and major pharmaceutical companies seeking to capture this market may face the risk of having to significantly revise their timelines for commercializing new drug candidates due to the FDA's slower decision-making process. Ultimately, the loss of regulatory flexibility could delay drug approvals, leading to reduced access to treatments for patients.