The US Food and Drug Administration has begun reviewing clinical-trial data in real time using AI, starting with two oncology trials. AstraZeneca's Phase 2 study of a combination therapy for an aggressive lymphoma is running at MD Anderson and the University of Pennsylvania; Amgen's Phase 1b trial covers a small-cell lung carcinoma treatment.

In parallel, the agency has issued a Request for Information on a wider "AI-Enabled Optimization of Early-Phase Clinical Trials" pilot, published in the Federal Register on 29 April 2026. The proposed scope covers safety monitoring, dose selection, and earlier go/no-go decisions in Phase 1 and 1b studies, all of which are normally constrained by how fast trial data can be cleaned and analysed by humans.

Two questions remain open. First, the announcement says nothing about which AI tools the FDA reviewers themselves will use, and the agency has been cautious in the past about model-specific endorsements. Second, real-time review by the regulator is a different proposition from real-time analysis by the sponsor, and the line between the regulator seeing sponsor data earlier and the regulator influencing trial conduct is where disputes are likely to arise.