FDA Seeks Input on Real-World Performance of AI-Enabled Medical Devices: What Biotech and MedTech Innovators Need to Know

On September 30, 2025, the Food and Drug Administration (FDA) issued a Request for Public Comment (“Request”) inviting stakeholders to share practical experience and recommendations for measuring and evaluating the real-world performance of AI-enabled medical devices, including those powered by GenAI. According to FDA, the Request is not intended to communicate new guidance or regulatory expectations but aims to advance the conversation on how best to assure the safety, effectiveness, and reliability of AI-driven technologies in clinical practice.

The Request outlines a series of targeted questions for stakeholders, including biotech and medical device companies. FDA specifically seeks input approaches for detecting and responding to performance changes, ensuring data quality, and integrating human oversight in the deployment of AI-driven medical technologies.

Given the current environment, commenting on the Request is an opportunity for companies to engage FDA on this important topic. For example, a company developing an AI-enabled diagnostic imaging tool might respond by describing its use of sensitivity, specificity, and false positive rates as performance metrics, monitored through integration with hospital EHR systems. They could explain a hybrid approach combining automated alerts for data drift with periodic radiologist review, and detail how user feedback from clinicians is incorporated into model updates. Similarly, a medical device manufacturer deploying a wearable cardiac monitor could discuss protocols for detecting abnormal input patterns, the use of patient-reported outcomes, and the role of user training in maintaining device reliability. FDA encourages submitters to share practical challenges, best practices, and privacy safeguards, helping to shape future regulatory approaches for AI-enabled technologies.

Hospitals and other covered entities are increasingly central to supplying the data necessary for AI-enabled medical devices, which raises critical issues around patient consent, authorizations, HIPAA compliance, and licensing. It is important for organizations and their partners to review consent forms and authorizations to ensure they are comprehensive and clearly address both initial and secondary uses of patient information and protected health information, including for device development and monitoring. In addition to federal requirements, organizations must remain attentive to evolving state AI, privacy, and cybersecurity laws, which may impose additional restrictions on the transfer, collection, use, and storage of personal information. Cybersecurity concerns also present significant implementation challenges, requiring robust safeguards for secure data transmission and integration with device platforms. The Request provides data supplying stakeholders with an opportunity to discuss the benefits and challenges in gathering, obtaining, and using their data for AI-enabled device evidence, helping to shape and guide FDA’s thinking in order to ensure appropriate patient protections are required and maintained.

As FDA’s approach to AI-enabled device oversight continues to evolve, proactive engagement now can help shape a regulatory framework that supports innovation while safeguarding patient outcomes.

Crowell and Crowell Health Solutions will continue to monitor for further guidance. Please reach out if you would like assistance in providing any comments to FDA’s request.