National AI Commission recommendations – what they mean for clinical radiology and clinical oncology

The National AI Commission published their recommendations on the future of AI regulation in healthcare on 10 September 2026. This is a fast-changing area for the health system, and with our specialties leading the way in NHS AI deployment, any recommendations taken forward by government could have a big impact on clinical practice.

We look at what the recommendations could mean for clinical radiology and clinical oncology, and how the College will continue to make sure our Fellows’ and members’ views and experiences continue to shape future policy and regulation.

Clinical radiologists and clinical oncologists are leading the adoption of AI in the NHS. Our 2025 workforce census found 75% of clinical radiology departments and 83% of cancer centres are using AI tools. Our Fellows and members are in step with the public - 78% support the use of AI within their specialities, while 80% of the public also supports the use of AI in radiology, as long as there is sufficient clinical oversight. Clinical oversight is key, both for maintaining patient safety and strengthening public trust in AI use in their healthcare. Indeed, as the Commission puts it in their report, the challenge is to develop a system that allows for “AI to do what AI is good at, and humans to so what they are good at” – in the case of clinical radiologists and clinical oncologists, that’s providing safe and effective care to all of our patients.

Improving regulation is critical if we’re to fulfil the potential of AI while maintaining trust. This is something we’ve been working on for some time, dating back to over a year ago when we held a workshop to inform our own position on AI Regulation which we published at the start of the year. This position formed the basis of our engagement with the AI Commission, throughout which we pushed for recommendations that address both the opportunities and concerns around AI that our Fellows and members have raised with us.

The National AI Commission was formed by the Medicines and Healthcare Products Regulatory Agency (MHRA) as an independent group with the task of advising the government on future AI regulation in healthcare. The Commission brought together representatives from across the health system to inform their recommendations.

We’ve been closely involved in this work, represented by our President and Vice-President for Clinical Radiology on the Health Systems Working Group advising the Commission. We also submitted written evidence which was developed in close collaboration with our AI Committee members and informed by the experiences of our Fellow and members.

The Commission’s recommendations suggest significant changes which, if implemented, will have wide-ranging implications for patients, clinicians, regulators and the AI industry.

1. The Commission recommends regulation throughout the product lifecycle, not just pre-market approval.

We’ve been pushing for this, because the performance of AI tools can change over time as systems interact with clinical realities. Earlier in the year we published post-deployment monitoring guidance and we’re working with clinicians currently using AI tools, developers and suppliers to identify what a national post-deployment monitoring system should look like. We’ll publish this work early next year.

There will also be an ‘L-plate’ system to allow for new tools to be trialled in a phased way, with close supervision.

2. The Commission recommends a shared responsibility for AI safety – across health systems, regulators, developers and manufacturers.

Under the proposals, the MHRA will have increased power to sanction developers of products, if they fail to meet required standards.

Clarity on responsibilities for AI safety would be a welcome step and reassuring for clinicians using the technology. However, the Commission didn’t make a clear call on medical liability. We'd like to do further work on this, as it’s a key issue for our members.

3. We need much more transparency on healthcare AI

The Commission recommends that patients should have the right to know whether AI is involved in their care, there should be a public database of adverse incidents involving AI-enabled devices, and a clear process for escalation if the performance of an AI tool is deteriorating.

We pushed for this clear escalation process and, if the government accepts this recommendation, we’ll be advocating for extensive clinician involvement in its development, so that the process is genuinely simple to use.

The government hasn’t formally responded to the Commission’s recommendations yet. When they do, and once we know if, how and by whom the recommendations will be taken forward, we’ll make sure we’re able to shape how this work is implemented.

We are pleased to see such ambitious recommendations, and a lot of the proposals align well with what our Fellows and members have told us is important – prioritising patient safety, more transparency and better, more consistent post-deployment monitoring. However, delivering this well will mean the MHRA, AI developers and services using AI will need to put more resource in – particularly in terms of monitoring. For some clinicians, building confidence and knowledge on AI will need to become a priority. With services already so stretched, there’s a risk that without support and resourcing to manage this workload AI adoption will be more challenging.

This is where we come in. We’ve already published guidance on AI deployment and post-deployment monitoring, and we’re currently working on proposals for a national post deployment-monitoring system. We run AI courses, and we have now been commissioned by NHS England to develop a national lecture series on AI, for all specialties. As well as this direct support, we will continue to work with our Fellows and members to shape the national approach, through our policy and influencing.

AI holds enormous potential for patient care and clinicians, but only if we get regulation and workforce support right. We will continue to champion the views of our Fellows and members and provide practical support, as our specialties, and the wider healthcare system, navigate these opportunities and challenges.