Nature urges real-world testing for AI-powered medical devices

September 29, 2026 at 07:58 PM1 min

This digest was compiled by AI from multiple sources — links to the originals are below.

Nature's editorial board calls for rigorous real-world testing of AI-powered medical devices, comparable to drug and self-driving car approvals. The US FDA is seeking feedback on regulating generative-AI-enabled medical devices, with a submission deadline of 19 October. The call comes as over 230 million people weekly ask ChatGPT health-related questions.

Key Facts

The US FDA published a discussion paper in August on regulating generative-AI-enabled medical devices.
The deadline for submissions to the FDA is 19 October.
More than 230 million people worldwide ask ChatGPT health-related questions every week.
Roughly three peer-reviewed articles about AI tools used in clinical medicine are published every day.

Regulatory Gaps

The FDA and its counterparts worldwide authorize medical products according to a scale of risk. All products must meet legally binding quality and safety standards, but not all need independent testing or real-world evaluation. Manufacturers of bandages can self-certify compliance, while makers of stethoscopes often need third-party authorization that can be done in a lab. Nature's editorial argues that AI tools informing clinical decisions require assessments equivalent to those for drugs and self-driving cars.

AI Adoption in Healthcare

Since the November 2022 release of OpenAI's ChatGPT, generative AI tools have entered clinics at speed and scale. AI systems can now handle complex administrative tasks, order laboratory tests, help prescribe drugs, and interpret X-rays, MRI scans, and CT images. They can also be used to diagnose rare diseases. In less than four years, thousands of studies have been published describing, testing, and evaluating AI use in healthcare.