FDA's AI Tool for Drug Approval Faces Accuracy Concerns

The Food and Drug Administration's artificial intelligence system designed to accelerate drug and medical device approvals is encountering significant operational problems, according to current agency employees.

The AI tool, called "ELSA," was introduced earlier this year with the goal of streamlining the approval process for pharmaceuticals and medical devices. However, three FDA employees report that the system is generating references to studies that do not exist, undermining confidence in its reliability for critical regulatory work.

Additional technical limitations include the AI's inability to access many relevant documents necessary for evaluating the safety and effectiveness of drugs and medical devices under review. These access restrictions further limit the tool's utility in the approval process.

The concerns arise as artificial intelligence applications in medicine operate with minimal federal oversight. FDA Commissioner Marty Makary stated this week that agency staff are not required to use AI tools in their work.

The Trump administration has identified AI as a potential method for reducing government spending in the health sector.