A UK commission set up to examine how artificial intelligence is regulated in healthcare has proposed wide-ranging changes for the NHS, including tighter oversight when new systems are first introduced, continuous monitoring after approval and clearer information for patients.
The recommendations come from the National Commission into the Regulation of AI in Healthcare, an independent body established by the Medicines and Healthcare products Regulatory Agency (MHRA) in September 2025. After a year-long investigation involving more than 12,000 patients, members of the public, clinicians, healthcare leaders, technology developers and industry representatives, the Commission said the current approach needs to better reflect how AI systems change over time.
Its report says people broadly support the use of AI in healthcare, but want strong safeguards in place. According to the findings, the public expects robust safety standards, meaningful human oversight and greater transparency about when AI is used in their care.
Professor Alastair Denniston, NHS consultant and Chair of the National Commission into the Regulation of AI in Healthcare, commented: "We innovate because we believe tomorrow’s healthcare can be better than today’s.
"But any new innovation – including AI – needs to earn its place and earn our trust. These technologies need to show that they are safe, effective and bring benefits to patients, staff and the wider NHS, without leaving people behind.
"Over the past year we have heard from more than 12,000 people, including patients, carers, clinicians and the technologists building AI tools. One message came through clearly: people are open to AI improving their care, but only if it is safe, overseen by humans, and if they know when it is being used.
These recommendations are designed to make that trust possible, so that the UK can lead the world not just in developing AI for healthcare, but in showing how it can be regulated and used responsibly.
Staged authorisation for new AI tools
One of the Commission’s main proposals is a staged authorisation model for new AI systems. Rather than moving straight from development into unrestricted clinical use, new tools would first be used under tighter controls, closer supervision and clearly defined safeguards.
The Commission compares the idea to restrictions placed on learner drivers. Under the proposed approach, safety and performance would be assessed in real-world healthcare settings before wider authorisation is granted.
It says this could help address two risks identified in the report: delaying access to potentially useful AI technologies for too long, and deploying systems too early before they have been properly tested.
Oversight beyond initial approval
The report also calls for regulation to continue throughout the working life of AI-enabled medical devices, rather than relying mainly on evidence gathered before they enter NHS use.
The Commission says this would be particularly relevant as AI systems are updated, adapted or exposed to new real-world data. Ongoing monitoring, it adds, would give the MHRA greater visibility over how products perform in practice and allow possible problems to be identified and addressed more quickly.
That marks a shift from treating approval as a one-off event to treating AI as a form of technology that may need continuing scrutiny.
More transparency for patients
The Commission’s recommendations also focus on how much information patients should receive. It says the public should have access to information about individual AI-enabled medical devices, including safety records and any reported adverse incidents.
It also found that people want to know when AI is involved in their healthcare. The report therefore recommends a proportionate approach to informing patients as AI becomes more integrated into NHS services.
AI is already used in the NHS for clinical and administrative work. Examples cited in the report include systems that help identify serious conditions such as strokes and skin cancers at an early stage, as well as AI-powered voice tools designed to reduce administrative workloads and give clinicians more time to focus on patients.
Call for stronger MHRA powers
The Commission has also recommended stronger enforcement powers for the MHRA. It says this would allow regulators to intervene more decisively when AI systems do not meet expected safety or performance standards.
The government and the MHRA will now consider the recommendations before issuing a formal response. The outcome could shape how the UK addresses what the report describes as a central challenge in healthcare AI: speeding up access to new technology while protecting patients when it is used in the consulting room.