MHRA publishes public views to shape AI healthcare rules

TL;DR:

• The MHRA has published two reports capturing how patients, clinicians and industry view AI in UK healthcare, drawn from one of the largest evidence-gathering exercises to date.
• A Call for Evidence drew 760 contributions, with broad agreement that existing regulation must adapt to AI’s pace.
• The findings feed the National Commission into the Regulation of AI in Healthcare, whose recommendations are expected later this summer.

The UK’s medicines regulator has set out what the public actually wants from AI in healthcare, publishing two reports that draw on patients, frontline staff, industry and policy experts. The exercise underpins the independent National Commission into the Regulation of AI in Healthcare and offers the clearest read yet on where consent and concern sit.

Cautious support, conditional on safeguards

The headline message is qualified enthusiasm. Contributors recognised AI’s potential to improve speed, quality and convenience of care, but only if rules guarantee safety and efficacy. A recurring theme was ongoing monitoring — checking that tools keep performing safely once in clinical use, not just at approval. Transparency, accountability and human oversight featured throughout, alongside a desire to move quickly enough that patients see benefits sooner.

A Call for Evidence gathered 760 responses from individuals and institutions, supplemented by deliberative sessions run with the Health Foundation and outreach to under-represented groups via the charity coalition National Voices. Professor Alastair Denniston, who chairs the Commission, said the input “will directly influence the final recommendations we make.” MHRA chief executive Lawrence Tallon framed the task as ensuring patients “benefit from innovation while remaining appropriately protected.”

The publication is the second MHRA AI move this month, following its sandbox to test AI in medicines development — a sign of how fast the regulator is building its position. It also lands amid live questions about liability, after warnings that doctors could be sued for AI mistakes.

Looking forward

The real test comes when the Commission reports this summer. These documents establish the evidence base; the recommendations will decide whether the UK adapts existing frameworks or builds new ones for clinical AI. For NHS leaders and health-tech firms, that verdict will shape what can be deployed, and how quickly, for years.