WASHINGTON — Artificial intelligence is moving deeper into American healthcare, and federal regulators are now trying to answer a difficult question: how can the United States speed up medical innovation without sacrificing patient safety?
The Food and Drug Administration is entering a new phase of digital-health oversight, with the agency developing regulatory approaches for artificial intelligence, generative AI, digital health technologies and software used in clinical care.
The shift comes as healthcare technology moves far faster than traditional regulatory systems were designed to handle. From AI-powered medical devices and wearable sensors to software that supports clinical decisions, digital tools are increasingly becoming part of diagnosis, treatment and drug development.
In August, the FDA took another significant step by releasing a discussion paper seeking public input on how generative AI-enabled medical devices should be evaluated before and after they reach patients. The agency is asking manufacturers, clinicians, researchers and the public to weigh in on risk assessment, premarket evaluation and postmarket monitoring.
FDA Opens New Debate Over Generative AI
Generative AI presents a different regulatory challenge from traditional medical software.
An algorithm designed for one narrowly defined task can often be tested against established performance standards. Generative AI systems, however, can produce different outputs depending on the data, prompts and circumstances involved.
That creates new questions for regulators.
How should the FDA determine whether an AI system is safe? How should regulators evaluate a system that can change over time? And how should performance be monitored after a product enters the healthcare market?
The FDA’s August discussion paper does not establish new regulations. Instead, the agency is seeking early feedback before developing a potential regulatory framework. Comments are due by October 19, 2026.
A Two-Part Risk Framework Is Being Considered
The FDA discussion paper outlines a possible two-axis approach to assessing risk for generative-AI medical devices.
The idea is to consider the characteristics of the technology and the potential consequences of its use.
The agency is also considering a competency-based approach to premarket evaluation. That could involve testing a system’s performance through nonclinical benchmarking followed by clinical confirmation.
The goal would be to determine whether an AI-enabled device performs as intended before it is widely used with patients.
Postmarket Monitoring Could Become More Important
The regulatory process may not end when an AI-enabled device receives authorization.
Generative AI can introduce risks that evolve after deployment.
That means regulators may need better systems for monitoring performance once products are being used in real-world healthcare environments.
The FDA’s discussion paper specifically raises questions about risk-proportionate postmarket monitoring, foundation models and agentic AI systems.
For technology companies, that could mean greater responsibility for tracking how their products perform after commercialization.
FDA Is Already Regulating AI-Enabled Medical Devices
The new discussion around generative AI does not mean the FDA is starting from zero.
The agency already maintains a list of AI-enabled medical devices authorized for marketing in the United States.
The FDA says these products have met applicable premarket requirements, including reviews addressing safety and effectiveness and whether the supporting studies were appropriate for the intended use and technological characteristics of the device.
The growing list demonstrates how quickly AI has moved from an experimental technology into mainstream medical-device development.
Digital Health Extends Far Beyond AI
The FDA’s digital-health responsibilities also include technologies that do not necessarily rely on artificial intelligence.
Digital health technologies can include software, sensors, connected devices and computing platforms used to collect or analyze health information.
Some technologies qualify as medical devices, while others do not.
The FDA’s Digital Health Center of Excellence serves as a resource for developers and other stakeholders working to understand the regulatory status of digital products.
Wearables Are Changing Clinical Research
One of the biggest changes is happening outside hospitals and traditional research facilities.
Wearable devices can collect health information continuously while people go about their daily lives.
Smartwatches, sensors, smartphones and other connected technologies can potentially measure activity, heart rate, glucose levels and other health indicators.
The FDA says digital health technologies can allow researchers to gather information remotely and continuously, potentially creating new ways to evaluate medical treatments.
That could make clinical trials more convenient for participants while generating data that traditional clinic visits may miss.
FDA Pushes Digital Tools Into Drug Development
The agency is also working to integrate digital technologies into pharmaceutical research.
In July, the FDA announced a funding opportunity focused on exploring how digital health technologies can be used in drug development.
The projects can examine digital measurements, new clinical-trial endpoints, continuous measurements and ways to identify early signs of chronic diseases.
The agency’s broader program includes a framework for digital health technologies, a steering committee, public meetings and demonstration projects.
A New Focus on Digitally Derived Measures
The FDA recently published a paper outlining key considerations for digitally derived measures used in clinical investigations.
The work involves multiple FDA centers and focuses on how digital measurements can be developed and used as outcomes in clinical research.
The agency says technologies enabled by AI can capture health information continuously and in real time, including outside traditional healthcare settings.
That could eventually change how clinical trials are designed.
Instead of measuring a patient’s condition only during periodic appointments, researchers may increasingly be able to observe health changes continuously.
Clinical Decision Software Gets New Guidance
Another major development came in January, when the FDA finalized updated guidance covering clinical decision support software.
The guidance clarifies which software functions may fall outside the legal definition of a medical device and which functions remain subject to FDA digital-health policies.
That distinction matters for developers.
Some software can provide information to healthcare professionals without becoming a regulated medical device, while software that performs certain medical functions may remain subject to FDA oversight.
The Agency Is Trying to Create Clearer Rules
The rapid growth of digital health has created a regulatory challenge for both the government and industry.
Technology companies need to know what evidence they must provide.
Healthcare providers need to understand whether digital tools are reliable.
Patients need confidence that technology used in their care has been evaluated appropriately.
The FDA’s evolving guidance is intended to provide more clarity while allowing innovation to continue.
A Faster Regulatory Model Is Emerging
The agency’s recent initiatives suggest a shift toward more flexible approaches for certain digital-health technologies.
One example is the TEMPO — Technology-Enabled Meaningful Patient Outcomes — pilot program.
The FDA announced its first TEMPO participant in July. The program is designed to support digital health devices intended to improve outcomes for Americans managing chronic diseases.
For qualifying devices used through the CMS Innovation Center’s ACCESS model, the FDA intends to exercise enforcement discretion for certain requirements, including some premarket authorization and investigational-device requirements.
The approach reflects an attempt to gather real-world evidence while allowing promising technologies to reach patients under controlled conditions.
Innovation Versus Patient Safety
The central debate is unlikely to disappear.
Technology companies argue that overly complicated regulatory requirements can slow innovation and increase development costs.
Patient advocates and medical experts, meanwhile, want strong safeguards for technologies that influence diagnosis and treatment.
The FDA must balance both sides.
A digital-health product that reaches patients too slowly may delay useful innovation.
A product that reaches patients without adequate evidence could create risks that are difficult to reverse.
AI Could Change How Doctors Work
The implications extend beyond regulators and technology companies.
AI-enabled tools could eventually help physicians identify disease, analyze medical images, monitor patients and interpret large amounts of clinical data.
That could reduce administrative workloads and potentially help doctors spend more time with patients.
But doctors will still need to understand the limitations of AI systems.
A tool that produces an incorrect recommendation could have serious consequences if healthcare professionals accept its output without appropriate review.
Transparency Will Become Critical
As AI becomes more common in healthcare, transparency could become one of the most important regulatory issues.
Patients may want to know whether an AI system contributed to a diagnosis or treatment recommendation.
Doctors may need information about how a system was tested and what populations were represented in the data used to develop it.
Regulators will also need methods for identifying performance problems after deployment.
Cybersecurity Is Another Concern
Digital health creates another category of risk: cybersecurity.
Connected medical devices and healthcare platforms can collect highly sensitive information.
A security breach could expose personal health data or interfere with the operation of medical technology.
The FDA already includes cybersecurity among its digital-health regulatory guidance areas, reflecting the increasing importance of protecting connected medical devices.
The Business Impact Could Be Significant
The regulatory changes could influence investment across the healthcare technology industry.
Companies developing AI-enabled medical devices may need to build regulatory expertise earlier in the product-development process.
Investors will also be watching how quickly the FDA establishes predictable pathways for emerging technologies.
Clearer rules could encourage investment by reducing uncertainty.
At the same time, stronger evidence requirements could increase development costs for companies building higher-risk systems.
U.S. Leadership Is Also at Stake
The FDA’s approach has implications beyond the American healthcare market.
The United States remains one of the world’s largest markets for medical technology.
Regulatory standards established by the FDA can influence how companies design products for other countries.
The agency’s latest generative-AI discussion paper specifically describes the effort as part of supporting U.S. innovation and global leadership in the technology.
The Next Phase Will Depend on Public Feedback
The FDA is now asking stakeholders to help shape the future framework.
Manufacturers, physicians, researchers, patients and members of the public can submit comments on the generative-AI discussion paper through October 19.
The agency says the document is intended for discussion and does not represent draft or final guidance.
That means the final regulatory framework could look different from the approaches currently being considered.
What Comes Next
The FDA’s digital-health agenda is likely to remain focused on three major areas: encouraging innovation, establishing reliable evidence and protecting patients.
The agency will continue developing guidance for AI-enabled medical devices while expanding work on digital health technologies used in clinical trials.
It is also testing new approaches that could allow certain technologies to generate real-world evidence while reaching patients more efficiently.
The challenge will be ensuring that faster access does not come at the expense of safety.
The Bottom Line
Digital health is entering a new regulatory era in the United States as the FDA works to keep pace with artificial intelligence, connected devices, wearables and increasingly sophisticated medical software.
The agency’s August discussion paper on generative-AI medical devices represents one of its most significant recent efforts to address the technology’s unique risks. The FDA is examining possible approaches to risk assessment, premarket testing and postmarket monitoring while asking the public and industry to provide feedback.
At the same time, the agency is expanding its use of digital health technologies in drug development and clinical research. Its programs include a dedicated steering committee, demonstration projects and initiatives designed to encourage remote and continuous data collection.
The TEMPO pilot adds another layer to the strategy by testing whether certain digital health devices can reach patients through a more flexible regulatory approach while generating real-world evidence.
For the healthcare industry, the direction is becoming increasingly clear: digital technologies are no longer a side issue for regulators.
They are becoming a central part of how medicine is developed, tested and delivered.
The FDA’s challenge now is to build a regulatory system that moves quickly enough for the age of AI while remaining rigorous enough to protect the patients who will ultimately rely on these technologies.
Source angle: FDA guidance and announcements on generative AI-enabled medical devices, AI-enabled medical devices, digital health technologies, digitally derived clinical measures, clinical decision-support software and the TEMPO pilot.