Can AI evaluate new medicines? FDA and OpenAI think so
What's the story
The US Food and Drug Administration (FDA) is said to be in talks with OpenAI to use artificial intelligence (AI) to speed up drug approvals. Recently, FDA Commissioner Marty Makary highlighted the agency's first AI-assisted scientific review for a product, a major step toward modernization. He stressed that this is just the beginning of integrating AI into their operations.
Partnership details
OpenAI's collaboration with FDA
While the exact details of the partnership remain under wraps, Wired reports that a small OpenAI team has met with the FDA and two associates from Elon Musk's Department of Government Efficiency on several occasions. The talks centered around a project called cderGPT, which stands for Center for Drug Evaluation and Research GPT. The discussions have been led by Jeremy Walsh, the FDA's first-ever AI officer, who was appointed recently.
Timeline reduction
AI's role in drug development timeline
Former FDA chief Robert Califf stressed that employing AI to help with final drug reviews could cut down a fraction of the lengthy drug-development timeline. Rafael Rosengarten, CEO of Genialis and cofounder of the Alliance for AI in Healthcare, backs automating some aspects of the drug-review process but calls for policy guidance around data used to train AI models.
Current procedures
FDA's existing mechanisms to expedite review process
The FDA's review process currently takes roughly a year, but there are mechanisms to expedite that timeline for promising drugs. These include the fast track designation for products designed to treat serious conditions and fill unmet medical needs, and the breakthrough therapy designation introduced in 2012. Andrew Powaleny from industry group PhRMA emphasized that while AI is still developing, harnessing it requires a thoughtful approach with patients at the center.