The U.S. Food and Drug Administration (FDA) has been engaging with OpenAI to explore how artificial intelligence (AI) can speed up the drug approval process. FDA Commissioner Marty Makary recently voiced concerns about the long timelines for drug approvals, noting that the integration of AI could be the solution to expedite these processes. In a post on X, he expressed the agency’s excitement about its first AI-assisted scientific review for a product, marking the beginning of AI’s role in streamlining the drug evaluation process, Wired reported.
The FDA is considering using AI for various parts of its operations, but sources close to the discussions reveal that OpenAI has been involved in early meetings with the agency. One of the projects under discussion is called cderGPT, which focuses on the Center for Drug Evaluation and Research (CDER). This AI-driven initiative would enable the FDA to enhance its review process, particularly for drugs related to conditions like diabetes and cancer. While the FDA has previously used AI in limited roles, expanding its usage could bring substantial improvements to the review timeline.
Walsh has also met with Peter Bowman-Davis, the acting chief AI officer at the Department of Health and Human Services, to further discuss the FDA’s AI initiatives. Bowman-Davis, an undergraduate on leave from Yale, is part of Andreessen Horowitz’s American Dynamism team, and his involvement in the FDA’s AI efforts underscores the growing influence of AI expertise in shaping the agency’s future strategies. This collaboration highlights the broader efforts to incorporate advanced technology in the FDA’s regulatory framework.
Despite enthusiasm from the FDA, experts like Rafael Rosengarten, CEO of Genialis, caution about the potential risks. He suggests that AI should first be tested on simpler tasks, such as checking application completeness, before moving on to more sophisticated uses like analyzing clinical trial data. He also pointed out that AI systems must be trained properly to ensure they produce accurate results. There are concerns about the potential for AI to fabricate or misinterpret information, which could lead to unsafe conclusions in drug approvals.
SEE MORE: Amazon is developing ‘Kiro,’ an AI-powered tool to simplify software coding
The FDA has mechanisms in place to fast-track drug approvals, such as the fast-track designation and breakthrough therapy designation. However, AI’s role would likely be focused on reducing review times for drugs already in the pipeline, rather than completely revolutionising the process. As OpenAI’s ChatGPT Gov, a government-compliant version of its AI, gets ready to comply with regulations and secure sensitive data, AI’s potential to reshape the future of drug approvals seems vast, but needs to be approached carefully.With the FDA’s ongoing efforts to modernise its internal operations with AI, the collaboration with OpenAI highlights the promising future of AI in regulatory science. By speeding up drug reviews, the FDA hopes to improve access to life-saving treatments while ensuring that safety and effectiveness standards are maintained. The next step will involve further testing and refinement of AI tools, with the goal of creating a more efficient, safer, and transparent approval system for the future.
