Public support for NHS AI drops sharply when diagnosis is involved
Three-quarters of the public would use the NHS App to book appointments or access procedure information. Support collapses to 49% when an AI system offers medical advice.
The Health Foundation's third annual health tech tracker surveyed 8,000 members of the public and 2,000 NHS staff. The results reveal a sharp divide between acceptance of administrative functions and wariness about AI making clinical decisions.
Booking hospital appointments drew 76% support. Choosing a preferred hospital: 73%. Accessing procedure information: 73%. The proposed AI-powered Doctor in Your Pocket feature for non-urgent care dropped to 49%, with 32% saying they would not use it.
The government signalled major plans to expand the NHS App as part of its 10 Year Health Plan. But the polling shows public trust remains the obstacle.
"The public currently prioritises stronger diligence and safeguards over potential benefits such as speed or availability," said Ahmed Binesmael, senior improvement analyst at the Health Foundation.
Healthcare leaders acknowledge the trust problem
Matthew Taylor, interim chief executive of the NHS Confederation and NHS Providers, said AI and digital tools have genuine potential to improve NHS productivity and give people control over their information. But he acknowledged the need to "build trust" in their use.
This mirrors concerns from professional bodies. The Royal College of Radiologists, the Institute of Physics and Engineering in Medicine, and the Society of Radiographers all called for properly trained staff and consistent regulation across AI developers and healthcare providers.
"AI must be regulated as a safety-critical technology," said Mark Knight, president of the Institute of Physics and Engineering in Medicine. "That requires clear standards across the AI lifecycle and a workforce with the capability and authority to assure these systems in clinical practice."
Legal framework lags behind technology
A separate concern: the law has not kept pace with AI's use in healthcare. Patients injured by faulty AI systems currently pursue product liability claims, which are complex, costly, and lengthy.
"People injured by medical products that utilise AI are forced to take on well-resourced manufacturers, sometimes based abroad," said Pauline Roberts, vice president of the Association of Personal Injury Lawyers.
APIL submitted evidence to the Medicines and Healthcare products Regulatory Agency as it develops a new regulatory framework for AI in healthcare. The consultation has closed, but no timeline exists for when recommendations will be published.
Government invests £1.6 billion in AI research
The UK government committed £1.6 billion to AI funding over four years through UK Research and Innovation (UKRI), its largest single investment area for 2026 to 2030. The strategy includes £137 million for AI-enabled scientific discovery, starting with drug discovery and new treatments.
The funding also supports new doctoral and fellowship routes, and career frameworks for research software engineers, data scientists, and ethics specialists.
Learn more about AI for Healthcare and AI for Science & Research.
Your membership also unlocks:
