US FDA to Accelerate Drug Trials With AI and Real-Time Data

FDA is launching a transition to real-time clinical trials: the regulator wants to see key safety and efficacy signals during the study itself, not after it concludes. If the scheme works, new drug development could be shortened by months, and in some cases years.

How the process is changing

Currently, data from clinical trials typically goes through a long chain: from research sites to the sponsor, then to analysis, and only after that to the FDA. In early stages this is particularly painful: patient numbers are low, uncertainty is high, and decisions about dosing, safety, and program continuation are often made with significant delays.

The new approach should reduce precisely this lag. The regulator wants to receive not the entire array of raw patient data, but key signals on efficacy and safety almost immediately, while the study is still ongoing. To achieve this, on April 28, 2026, FDA announced two immediate steps. First — launching proof-of-concept trials in which signals will be transmitted to the regulator in real time. Second — issuing a request for proposals for a broader pilot for early-phase studies.

In a document published on April 29, the agency separately emphasizes that the pilot will be built on NIST principles for reliable AI systems: validity, safety, security, accountability, explainability, privacy protection, and fairness. Comments on the program are accepted until May 29, 2026.

Where to begin

In the first phase, FDA is working with two oncology studies. AstraZeneca is already conducting a multicenter Phase 2 TRAVERSE study for patients with previously untreated mantle cell lymphoma in collaboration with MD Anderson and the University of Pennsylvania. Amgen is launching a Phase 1b STREAM-SCLC study for patients with limited-stage small cell lung cancer.

For AstraZeneca's project, the agency has already confirmed the technical feasibility of receiving and validating signals in real time through the Paradigm Health platform — this became the main proof that the scheme is actually applicable outside presentations.

• Patient recruitment for the study
• Dose selection and escalation
• Safety signal monitoring
• Adaptive protocol design
• Earlier go/no-go decisions

The next step is a summer pilot with a broader circle of sponsors. FDA wants to understand where AI truly provides benefits: in first-in-human studies, in oncology, in rare diseases, or in tasks like recruitment and biomarker selection. The agency is also gathering input on how to compare the AI-based approach with the traditional process, what metrics should be considered success, and what infrastructure needs to be built for safe data sharing between companies, research centers, and the regulator.

Where the benefit and risk are

The potential gain here is not only that drugs advance to the next phase faster. If the regulator sees critical signals immediately, dosing can be adjusted earlier, safety can be assessed more accurately, and decisions about continuing or stopping a program can be made faster. According to FDA leadership, the new model should challenge the old assumption that bringing a drug to market inevitably stretches to 10–12 years.

For companies, it's also a chance to eliminate some of the administrative 'dead time' that doesn't create scientific value but delays development.

«For sixty years we've been conducting clinical trials almost the same

way, and important signals could take years to reach the FDA.»

But along with speed come higher demands on system quality. The regulator directly asks the market how to measure model accuracy, track drift, verify result robustness across different sites, and maintain control over data privacy and integrity. A separate issue is explainability: if AI recommends accelerating the transition between phases or helps filter out patients, participants in the process must understand why the system came to that conclusion.

So far, the discussion is not about automatic drug approval, but about earlier, more intensive, and better-measured analytics.

What this means

If the pilot confirms expectations, FDA will begin to change its own role: from a regulator that sees the study after the fact, it will transform into a participant in more continuous monitoring. For pharma, this is a chance to cut months off the bureaucratic cycle without formally loosening requirements. For patients — the opportunity to access promising therapies faster, especially in oncology and rare diseases. But the main test lies ahead: will AI truly accelerate clinical science without turning data quality and safety into collateral damage.