National Commission sets out recommendations for the future of AI in healthcare
The National Commission into the Regulation of AI in Healthcare has published its recommendations for how the UK’s regulatory framework should evolve as AI becomes increasingly used across healthcare.
The recommendations were developed under the leadership of Professor Alastair Denniston (INSIGHT emeritus director) who chairs the National AI Commission and Professor Henrietta Hughes, Deputy Chair of the Commission and Patient Safety Commissioner for England. “This is about creating a future healthcare system that is increasingly tech-enabled but always people-centred,” said Professors Denniston and Hughes.
Professor Pearse Keane, Director of INSIGHT at Moorfields Eye Hospital and Professor of Artificial Medical Intelligence at UCL, contributed as a member of the Working Group to the Commission. “We welcome and support the Commission’s recommendations, which align with our ongoing activity to inform the development and translation of safe, equitable and well-evidenced medical AI that works well for everyone,” said Professor Keane.
The Commission’s recommendations have been shaped by evidence, expertise and extensive engagement. This has included input from its four expert working groups and wider research and engagement with patients and the public, healthcare professionals, healthcare organisations, industry, researchers, regulators and others across the health system. The evidence showed strong support for the use of AI in healthcare, but that support is conditional rather than automatic. People want to know that AI can be used safely, fairly and transparently, with clear evidence of benefit, meaningful human oversight and accountability.
The Commission’s central conclusion is that the regulation and assurance of AI in healthcare needs to become more proportionate, lifecycle-based and system-wide. Its vision is for a future framework that is Safe, Fast and Trusted: Safe in protecting patients and managing risk; Fast in enabling responsible innovation and timely access to beneficial technologies; and Trusted in giving confidence across the healthcare system.
The Commission’s recommendations are organised around three connected areas of change.
1. Proportionate lifecycle regulation This means having the right safeguards, at the right time, for the right level of risk. Regulatory requirements should reflect what a technology does, the benefits it could bring, the risks it presents and the level of uncertainty around it. Safe, low-risk tools that support staff, improve services and patient care, should face lighter requirements, so those benefits are not unnecessarily delayed. However, products that have a greater effect on someone’s health or care, or where there is greater uncertainty, would need stronger safeguards matched to their level of risk. Oversight must continue throughout the lifecycle, this means ensuring safety from the start but also continuing through deployment, monitoring, updating and ongoing learning from real-world use.
2. System-wide responsibility and safe management This means recognising that keeping AI safe is a shared responsibility. Regulation alone cannot ensure the safe and effective use of AI. Manufacturers, healthcare providers, healthcare professionals, regulators and policymakers all have a role to play. The future framework should support clearer responsibilities, stronger organisational readiness, appropriate workforce capability and effective governance. No one should be left responsible for something they cannot understand, control or act on.
3. Trust, transparency and predictability This means recognising that trust must be earned. Patients and healthcare professionals need clear and accessible information about how AI is being used, alongside transparency, accountability and meaningful involvement so they can have confidence in the technologies supporting their care. Developers and manufacturers also need clearer and more predictable pathways through the regulatory system, helping them understand what is expected and giving them greater confidence to bring safe and beneficial technologies into healthcare responsibly.
The recommendations are intended to give people greater confidence that AI used in healthcare is safe, effective, fair and properly governed, while keeping people and their care at the centre.
