The FDA announced real-time clinical trials on April 28. Two proof-of-concept trials are live. The program dissolves seventy years of phase-gate infrastructure, and with it the venture financing, pharma licensing, and biotech trading structures that synchronized around discrete trial phases.
On April 28, the FDA announced its first-ever real-time clinical trials program. Two proof-of-concept trials are already running. A request for information is open through May 29 to recruit additional participants. Pilot selections are expected in August.
The announcement received modest coverage. It will restructure how drugs are developed, financed, licensed, and traded.
What Changed
For seventy years, drug development has operated on a batch-processing model. A company runs a trial phase, compiles data over months or years, submits the batch to the FDA, waits for review, then proceeds to the next phase. Phase 1, Phase 2, Phase 3. Each gate is a discrete submission-and-review cycle. The model was built for a paper-based regulator reviewing printed case report forms.
The real-time clinical trials program replaces batch review with continuous data streaming. Paradigm Health, the sole validated technology platform, captures data directly from electronic health records and unstructured clinical sources, algorithmically evaluates FDA-defined data points, and transmits critical signals in real time. Data that previously took months to compile and review is analyzed in days.
AstraZeneca is running a Phase 2 multi-site trial called TRAVERSE for treatment-naive mantle cell lymphoma across MD Anderson and the University of Pennsylvania. Amgen is running a Phase 1b trial called STREAM-SCLC for limited-stage small cell lung carcinoma with final site selection in process. These are the only two participants so far. The FDA has received and validated signals from AstraZeneca's trial, establishing technical feasibility.
What It Dissolves
Clinical trial phases were never properties of biology. They were properties of how long a paper-based regulator took to review batched submissions. The distinction between Phase 1 and Phase 2 is an administrative boundary, not a scientific one. Continuous data streaming makes the boundary unnecessary.
The downstream effects are structural. Venture capital in biotech has been milestone-funded by phase: seed money to reach Phase 1, Series A to reach Phase 2, Series B to reach Phase 3. Each financing round is priced against the probability of clearing the next gate. Remove the gates and the tranched financing model loses its milestones.
Pharmaceutical licensing deals are structured the same way. A large pharma company licenses a molecule from a biotech startup with milestone payments triggered at phase transitions: $50 million at Phase 2 entry, $200 million at Phase 3 entry, $500 million at approval. These deal structures assume discrete events. Continuous regulatory review produces continuous information instead of binary phase-change events.
Biotech trading on the buy side is organized around the catalyst calendar. Analysts track Phase 2 data readouts and Phase 3 topline results as binary events that move stock prices 30 to 80 percent in a session. Options strategies, event-driven hedge funds, and retail speculation all depend on the discrete-event structure. A continuous data flow produces continuous repricing instead of catalyst spikes.
Who Built It
Jeremy Walsh was appointed the FDA's inaugural Chief AI Officer in May 2025. He spent 14 years at Booz Allen Hamilton as chief technologist, building cloud infrastructure and data analytics for the FDA, CDC, NIH, the VA, and military health systems. He has over 20 years of experience in distributed software systems. He is a government IT modernizer, not a pharmaceutical executive.
This matters because the program was designed by someone who sees clinical trials as a data infrastructure problem, not a drug development problem. The solution is an engineering solution: stream the data, evaluate it algorithmically, transmit signals continuously. The regulatory apparatus remains. The batch-processing assumption is removed.
The International Gap
The European Medicines Agency issued its first AI methodology qualification opinion in March 2025 for an inflammatory liver disease diagnostic tool. The FDA and EMA published 10 joint AI principles in January 2026 covering lifecycle monitoring, model drift assessment, and dataset bias evaluation. The UK's MHRA runs an AI Airlock regulatory sandbox. Japan's PMDA is shifting to what it calls an incubation function for cutting-edge medical technology.
None of them have announced real-time trial data streaming. The FDA is the first major regulator to move from batch review to continuous review. The international gap creates a window in which companies that build infrastructure for continuous regulatory interaction have a structural advantage in the world's largest pharmaceutical market.
Winners and Losers
Paradigm Health is the sole platform that has been validated by the FDA for real-time signal transmission. Every additional trial participant in the pilot will need compatible infrastructure. Contract research organizations that adapt to continuous data flows will take share from those built on the phase-gate batch-processing model.
Traditional CROs built on discrete-phase batch processing face structural pressure. PPD and Syneos Health sell project-scoped engagements organized around phase transitions. A continuous data model changes what they sell from project delivery to infrastructure operation. Biotech venture firms whose fund structures assume milestone-based returns will need to reprice how they value clinical-stage companies when the milestones dissolve.
The RFI comment period closes May 29. Watch whether incumbents who benefit from the current gated model push back through the comment process. The pilot selections in August will determine how quickly continuous review scales beyond two trials.
The FDA just replaced a seventy-year-old filing cabinet with a data stream. The companies, funds, and traders who organized their businesses around the filing cabinet's rhythm have until August to decide what that means for them.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)