Skip to main content
Government report Authoritative — Peer-reviewed venues, standards bodies, regulators, official government publications

National Commission into the Regulation of AI in Healthcare: Recommendations for a future regulatory framework

Final report of the UK National Commission into the Regulation of AI in Healthcare, established in September 2025 to advise government on a regulatory framework for software and AI-enabled health technologies. Sets out 38 recommendations under three principles (proportionate lifecycle regulation; system-wide responsibility and safe management; trust, transparency and predictability), covering device qualification and classification, risk-proportionate evidence over the lifecycle, general-purpose AI models used inside medical devices, direct-to-consumer products, post-market assurance, adverse-incident sharing, enforcement, and patient transparency.

Publisher

Medicines and Healthcare products Regulatory Agency (MHRA), National Commission into the Regulation of AI in Healthcare

Published

10 Sept 2026

Added

today

DOI

Key Findings

  • Recommendation 7 proposes an opt-in 'master file' approach for general-purpose platforms, foundation models and underlying technologies used by device manufacturers, and Recommendation 8 asks manufacturers to report product dependencies on underlying general-purpose models; the report names general-purpose models fine-tuned to serve as the basis of a medical device, such as a chatbot, as the case in point.
  • Recommendation 11 asks the MHRA to provide guidance on regulatory expectations for direct-to-consumer applications and wearables, including when such a product is a medical device and evidence expectations, recognising that consumers cannot rely on a health professional for risk controls.
  • Recommendations 19 and 22 propose a database or tool for sharing information on adverse incidents and enhanced enforcement mechanisms for the MHRA; Recommendation 6 covers predetermined change control plans.
  • Recommendations 35, 37 and 38 cover patient transparency about AI in care (with opt-out where possible or appropriate), public communication on safety concerns, and an 'AI sentiment census' run by DHSC and the devolved health departments.
  • The report distinguishes lower-risk uses (administrative, general wellbeing, low-risk clinical decision support) from higher-risk uses, and frames the whole healthcare system rather than the regulator alone as responsible for safe deployment.

Methodology Notes

Policy commission chaired by Professor Alastair Denniston with Professor Henrietta Hughes (Patient Safety Commissioner for England) as deputy chair; evidence from a call for evidence, public deliberation, engagement with seldom-heard groups, roundtables and working groups across the four UK nations (the companion 'Research, engagement and call for evidence findings' publication of 11 June 2026 carries the polling). Recommendations document with no primary study data. Full report PDF 119 pages (created 8 September 2026), plus easy-read and Welsh summaries; published on GOV.UK 10 September 2026 under the Open Government Licence.

Tags

mhrauknational-commissionmedical-devicesdirect-to-consumergeneral-purpose-aiadverse-incidents

Cite This

APA

Medicines and Healthcare products Regulatory Agency (MHRA), National Commission into the Regulation of AI in Healthcare. (2026). National Commission into the Regulation of AI in Healthcare: Recommendations for a future regulatory framework. https://www.gov.uk/government/publications/national-commission-into-the-regulation-of-ai-in-healthcare-recommendations-for-a-future-regulatory-framework