Jump to content

UK healthcare AI commission recommendations (2026)

From AI Law Wiki

AI-generated text. This page was generated using artificial intelligence.

Law & policy information
Name
National Commission into the Regulation of AI in Healthcare — recommendations[1]
Jurisdiction
United Kingdom[1]
Instrument type
Advisory recommendations for regulatory reform[1]
Bill / legal citation
Commission report, published September 10, 2026[1]
Legislative or adoption status
Government accepted all 44 recommendations in its October 6, 2026 response and set initial delivery priorities; legislation and detailed guidance remain to be developed.[2]
Effective dates
Publication: September 10, 2026. The report does not itself enact its recommendations.[1]
Last source verification
October 6, 2026[2]

The National Commission into the Regulation of AI in Healthcare recommends a framework for oversight across the lifecycle of healthcare AI, with responsibilities extending beyond medical-device regulation. Its September 10, 2026 report advises government; it is not itself a new statute or binding regulatory instrument. The government accepted all 44 recommendations on October 6, 2026 and set initial priorities for implementation.[1][2]

Scope

The report addresses software and AI-enabled medical devices, alongside accountability, transparency, professional practice and organisational governance. It distinguishes medical-device regulation from other applicable frameworks, including data protection and professional standards; not every healthcare AI application is regulated as a medical device.[1]

The commission calls for proportionate oversight throughout development, deployment, updating and monitoring, with real-world evidence supporting ongoing assurance. It proposes clearer responsibilities for developers, manufacturers, healthcare organisations and professionals, better communication of safety concerns, and stronger patient and public involvement.[1]

Recommendation 14 proposes staged authorisations for suitable devices. Initial deployment would have a controlled scope, safety evidence, risk controls and reporting conditions. Authorisation could expand after specified evidence thresholds are met. The report calls for transparent pathways tailored to technical maturity, clinical needs and patient safety; it does not grant individual products authorisation.[1]

Implementation status

On October 6, 2026, the government accepted all 44 recommendations and committed to a lifecycle-based approach to AI-enabled medical-device oversight. The MHRA said it would issue draft guidance by December 2026 on managing changes to AI devices, consult in 2027 on device qualification and classification, and publish a full implementation roadmap by spring 2027. These are commitments to develop policy and guidance; the response itself does not amend the medical-device regulations.[2][3]

The MHRA opened phase 3 applications for its AI Airlock regulatory sandbox, concentrating on post-market surveillance. Applications remain open on a rolling basis, with October 31 as the cutoff for the current review wave.[4]

References