FDA to deploy AI assistant for research reviews by end of June

The Food and Drug Administration launched an “aggressive timeline” on Thursday that charts its internal efforts to incorporate artificial intelligence across all of the agency’s 11 centers by this summer for use in scientific research reviews. 

Following a pilot on generative AI tools’ efficacy in aiding scientific reviews, FDA Commissioner Martin Makary announced the deployment of the finalized version of the software across the agency’s specialized offices via a secured and unified platform.

The AI tool is expected to handle the more tedious, repetitive aspects of the FDA’s scientific review to ensure agency scientists and experts are free to focus on critical components of the safety evaluation process. 

“I was blown away by the success of our first AI-assisted scientific review pilot,” Makary said in the press release. “We need to value our scientists’ time and reduce the amount of non-productive busywork that has historically consumed much of the review process. The agency-wide deployment of these capabilities holds tremendous promise in accelerating the review time for new therapies.”

He further added that the official launch of an AI review assistant follows "years of talks about AI capabilities.” FDA centers are now asked to fully deploy the AI tool by June 30 and have it fully integrated into individual data platforms and networks.

The agency hopes to ultimately expand the tool’s use cases, further enhance its capabilities for document integration and refine its outputs to center-specific needs and work. The FDA’s new chief AI officer, Jeremy Walsh, and the Director of the Center for Drug Evaluation and Research’s Strategic Programs Office Sridhar Mantha will both be at the helm of the implementation efforts.

Greater transparency in the deployment process, such as user feedback and performance results, will also be publicly available in June. 

In a statement sent to Nextgov/FCW, an FDA spokesperson said that Makary’s posture on AI integration is to enhance human expertise without replacing it.

“When used responsibly, AI can enhance regulatory rigor by helping predict toxicities and adverse events for certain conditions,” the spokesperson said. 

The FDA has long eyed AI as a means of creating a more efficient drug and therapy evaluation process. In 2023, the agency worked across its offices to understand how to deploy AI-powered analytics for a wide variety of the FDA’s data processing needs. It also released a contract opportunity to procure AI-enabled imaging technology to leverage in drug product quality assessments the same year.

Alongside the agency’s AI modernization ambitions, the FDA also continues to maintain its webpage devoted to resources on potentially harmful algorithmic biases in AI and machine learning-powered medical systems.