AI in the NHS: public attitudes and policy priorities

We have published four reports: our position on AI regulation in healthcare, the findings of public polling on attitudes towards the use of AI in healthcare, a consensus statement which outlines the work still to be done in terms of the deployment of AI in diagnostics in the NHS, and the findings of a high-level roundtable discussion of global thought-leaders in healthcare AI at the RCR-NHS Global AI Conference 2025. 

Our position on AI regulation notes the growing, assistive role at AI is playing in our specialities. It highlights the need for a regulatory framework that strengthens’ clinician and, in turn, patient confidence in AI use and, in particular, recommends:

1. Establish a principle of data transparency and high levels of evidence which reflect the complexity of the task of the AI tool and/or the potential patient impact.
2. Develop a single, national post-deployment monitoring system for AI which provides clinicians with a simple means to flag errors or concerns and a fast-paced feedback loop.
3. Prioritise a national deliberation on medical liability in cases where AI has been used, involving all relevant parties, including industry, regulators, clinicians, Trusts/health boards, civil society and patients.

Our polling report shows what the public knows and feels about the use of AI in their healthcare. It reveals overwhelming support for doctors to lead and shape how AI is used, alongside a need for greater public education to build and protect trust. Some of the headline findings include: 

• Less than half (46%) of the public have heard of AI being used in the NHS 
• A clear majority (80%) supported the use of AI in radiology in some way, and a majority were convinced that AI could improve early detection and diagnosis of cancer. 
• The public are highly trusting of doctors and the NHS, especially with the use of their personal health data. 

Our consensus statement on AI use in diagnostics illustrates the importance of AI tools that can alleviate doctors’ administrative burden, as well as the crucial role played by data and the need to balance data access with regulation to protect sensitive information.  

Global approaches to AI in radiology examines some of the ways in which AI is currently being used in diagnostics in countries around the world. It identifies specific lessons the UK could learn from other jurisdictions such as Germany, Sweden, South Korea and others.