FDA Rejects Proposal to Ease Oversight of AI Medical Devices

The FDA on May 16, 2026 rejected a proposal from Australian health AI company Harrison.ai to partially exempt certain AI-enabled diagnostic software from traditional 510(k) premarket review requirements. Harrison.ai's proposal, which received 47 comments through February 27, 2026, sought reduced barriers for developers with previously authorized products. More than 1,000 FDA-authorized AI-enabled medical devices are already on the market, with radiology among the largest categories.

FDA rejection of Harrison.ai proposal

• ▪Harrison.ai asked the FDA to partially exempt several categories of AI-enabled diagnostic and detection software from traditional 510(k) premarket review requirements under specific conditions
• ▪Harrison.ai argued that developers with previously authorized products and strong post-market monitoring programs should face fewer regulatory barriers when introducing similar AI systems
• ▪The Harrison.ai proposal was published in the Federal Register and accepted feedback through February 27, 2026, during which 47 comments were received
• ▪The U.S. Food and Drug Administration rejected an industry proposal from Australian health AI company Harrison.ai that would have reduced premarket review requirements for certain AI-enabled medical devices

AI medical device oversight debate

• ▪The FDA recently issued guidance documents viewed by some observers as loosening oversight of certain lower-risk AI-powered clinical decision support and wellness technologies
• ▪Supporters of lighter oversight have warned that lengthy regulatory pathways could slow innovation and delay patient access to potentially beneficial technologies
• ▪Researchers, clinicians, and patient safety advocates have cautioned that AI systems capable of influencing diagnosis or treatment decisions require strong safeguards, especially when algorithms may produce errors or affect clinical decision-making
• ▪More than 1,000 FDA-authorized AI-enabled medical devices are already on the market, with radiology among the largest categories

Diagnostic AI regulatory requirements

• ▪FDA officials argued that prior authorization of one AI device does not necessarily demonstrate that future products developed by the same company will perform safely or effectively
• ▪The Harrison.ai proposal focused on several categories of radiology and diagnostic AI software, including computer-aided detection and diagnostic systems used to identify abnormalities in medical imaging
• ▪The FDA raised concerns about relying heavily on manufacturers' internal monitoring systems without sufficient FDA oversight
• ▪The FDA maintained existing oversight requirements for the affected AI device categories in its rejection of the Harrison.ai proposal

Global AI healthcare regulation

• ▪The European Union has adopted stricter, risk-based oversight under the EU AI Act, which classifies many medical AI systems as high-risk technologies subject to heightened compliance obligations
• ▪Differences in how countries regulate AI may lead companies to prioritize markets with fewer barriers, potentially affecting access to digital innovations and how safety and transparency protections are applied from the onset
• ▪The United States has generally pursued a more flexible, state-led regulatory model for AI in healthcare so far

Innovation versus patient safety balance

• ▪The FDA decision may indicate that regulators are increasingly distinguishing between lower-risk administrative or wellness-focused AI systems and AI tools that directly influence clinical decision-making
• ▪Policymakers are increasingly confronting questions surrounding clinical validation, transparency, algorithmic bias, post-marketing monitoring, and liability for AI-assisted decisions as AI-powered medical devices expand rapidly
• ▪The FDA decision on May 16, 2026 highlights the growing policy tension between promoting innovation and ensuring patient safety as AI adoption accelerates in healthcare

Perspective of Patient safety advocates and clinicians

• ▪Researchers, clinicians, and patient safety advocates have cautioned that AI systems capable of influencing diagnosis or treatment decisions require strong safeguards, especially when algorithms may produce errors or affect clinical decision-making