New research finds accuracy and oversight are public's top priorities for AI regulation in healthcare
The research, conducted by Ipsos for the Health Foundation, finds that the public sees significant potential benefits from AI in health but prioritises accuracy and oversight over faster adoption.
The research was designed to understand public priorities for a regulatory approach to Artificial Intelligence (AI) in health. It comprised three deliberative workshops, taking place online and in-person during March and April 2026, in York, Milton Keynes and Cardiff. In total 78 participants were included in the research, broadly reflecting the wider population living in the UK
The key findings from the research are:
- Participants recognised the potential for AI to improve health and care, particularly by reducing administrative burdens.
- Accuracy was the public's highest priority for AI regulation. Participants set a high bar for performance, generally expecting AI tools to be at least as good as the relevant clinician, given enough time in a real-world setting.
- Human oversight was seen as an important safeguard, particularly for higher-risk uses of AI (such as diagnosis). Participants generally viewed AI as a tool to support clinicians rather than replace them.
- Participants prioritised safety over speed - which was valued for tackling waits, but a secondary benefit unlocked only after accuracy and oversight are assured.
- The public supported a risk-based approach to regulation, favouring stricter requirements for higher-risk applications (such as AI being used in making diagnostic decisions), while accepting lighter requirements for lower-risk uses (such as administrative tasks), provided robust oversight remained in place for all uses.
- Participants wanted AI systems to be continuously monitored after deployment, with clear accountability and the ability for regulators to intervene if problems emerged.
