2MM: AI Roundup – FDA’s AI Push, Trial Speedups with Real-World Data, Smart Surgical Monitors, and Regulatory Overhaul Begins [May 23rd, 2025]

Food and Drug Administration to Fully Integrate Generative Artificial Intelligence by June
The Food and Drug Administration is diving headfirst into artificial intelligence, announcing plans to deploy a generative artificial intelligence system across all its centers by the end of June. Leading the charge is Jeremy Walsh, recently appointed as the agency’s first Chief Artificial Intelligence Officer. On May 15, officials shared that a pilot tool, called Center for Drug Evaluation and Research Generative Pretrained Transformer, is already in testing to assist with drug application reviews. The artificial intelligence aims to streamline operations and cut down the time required for regulatory approvals. In a promising sign, the Food and Drug Administration has already completed its first scientific review assisted by artificial intelligence. Walsh is also exploring potential collaborations with OpenAI to enhance the agency’s technological capabilities. The system could free staff from time-consuming paperwork and allow for faster decision making. If the timeline holds, this technology could be live in just a few weeks.

Artificial Intelligence and Real World Data Speed Up Drug Development
TriNetX is transforming how clinical trials are conducted by harnessing artificial intelligence and real world data from electronic health records. The company reported on May 10 that its platform is helping researchers identify eligible patients in real time while also improving diversity and design in trials. This has already led to a 40 percent drop in enrollment times for some studies. Their artificial intelligence models simulate disease progression, offering earlier insights into trial outcomes, especially in hard to study conditions like pancreatic cancer. A digital waiting room feature helps match patients to trials quickly, bypassing traditional recruitment delays. This innovation could mean major savings and faster access to new treatments. With biotechnology budgets under pressure, these tools are arriving just in time. Artificial intelligence is no longer an add on in drug development, it is becoming essential.

Becton Dickinson Unveils Artificial Intelligence Powered Monitor for Surgery
Becton Dickinson’s new surgical technology is turning heads for good reason. On May 13, the company introduced the HemoSphere Alta system, a surgical monitor powered by artificial intelligence that predicts drops in blood pressure during operations. It reads patient data in real time using noninvasive sensors and machine learning, offering surgeons early warnings that help reduce complications. During early trials, the device improved predictive accuracy by over 30 percent compared to traditional monitors. By keeping cerebral perfusion steady, it could improve recovery outcomes and lower the risk of intraoperative emergencies. Becton Dickinson sees this tool becoming a staple in operating rooms across the world. It is part of a broader trend of artificial intelligence stepping into direct patient care. The operating room is getting smarter, one algorithm at a time.

Artificial Intelligence Tool Matches Patients to Clinical Trials with Near-Human Accuracy
The National Institutes of Health is making clinical trial recruitment faster and smarter with a new tool called TrialGPT. Unveiled in May, the artificial intelligence model is designed to match patients to suitable clinical trials using data from ClinicalTrials.gov. In a recent study, TrialGPT achieved 87.3 percent accuracy, coming close to the performance of human reviewers. Clinicians using the tool were also able to screen patients 40 percent faster than when using manual methods. The system works by generating relevant keywords for each patient and producing short summaries that explain the match. While there were occasional errors in interpreting some medical procedures, the tool’s performance marks a major step forward in automating a slow and expensive process. Given that most clinical trials struggle with enrollment, this kind of technology could make a meaningful difference. The team behind TrialGPT is now testing it in real-world clinical settings to further refine its impact.

Image: PD

©2025 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.