A recent major study from King’s College London highlighted a rapid shift in how patients are engaging with AI, finding that one in seven (15 per cent) have used a chatbot for health advice instead of contacting a GP. The trend underlines growing consumer confidence in digital tools but also raises concerns about how these systems are shaping clinical decision‑making, particularly given that one in five users say AI discouraged them from seeking professional care.
Despite rising usage, the research shows public trust remains conditional. Three quarters (76 per cent) of respondents say AI used in patient care should be officially approved and regulated, even if that slows its rollout.
Graham Rapier, CEO of clinical AI video platform 4Cori, based in the Harley Street Health District, says the findings mirror what he hears from both clinicians and patients. He sees this as the natural evolution of any emerging technology, where adoption is shaped as much by user confidence as capability. He emphasises that as AI tools improve, accountability should stay firmly with clinicians.
In practice, he argues, this means AI tools should be built around clinical workflows rather than replace them. The likely future, he suggests, is “clinician‑started, AI‑accelerated and clinician‑approved,” akin to aviation, where sophisticated autopilot technology is in use, but the pilot remains in control.
Mr Asim Kazmi, Head of Digital and AI at Optegra, also based in the District, said the public’s split views on AI in clinical decision-making should not be seen as resistance, but as a demand for systems that are “regulated, transparent and always accountable to a clinician”.
He added that patients are already turning to AI because of difficulties accessing care, making the key question whether they use “a general-purpose chat bot” or a governed clinical pathway overseen by clinicians. In ophthalmology, he says AI can support earlier detection and faster triage for conditions such as diabetic retinopathy, glaucoma and macular degeneration, but stresses that it should strengthen, not replace, clinical judgement. “Scepticism isn’t the enemy of AI innovation,” he adds, “unearned trust is.”
