The Medicines and Healthcare merchandise Regulatory Company (MHRA) has suggested that AI chatbots shouldn’t substitute recommendation from healthcare professionals after analysis discovered one in 4 UK sufferers are turning to AI and social media for well being steering.
Semble, a startup that gives a present a personal healthcare platform, surveyed 2,000 UK sufferers reflecting a spread of backgrounds, ages and well being wants in August 2025.
The analysis discovered that 24% of sufferers use social media and AI instruments comparable to ChatGPT or Instagram for well being steering, and 30% could be prepared to seek the advice of AI or social media moderately than wait to see a clinician.
Gen Z are among the many most energetic adopters, with 34% of 16-25-year-olds turning to ChatGPT and 30% utilizing TikTok.
Responding to the outcomes, Dr Ed Middleton, director of technique on the MHRA, mentioned that the Nationwide Fee into the Regulation of AI in Healthcare, which launched in September 2025, will be certain that applied sciences utilized in healthcare are “evidence-based, clear and correctly regulated”.
“The ten 12 months well being plan set the ambition to make the NHS essentially the most AI-enabled well being system on the planet.
“The work of the fee will assist guarantee that innovation is matched with the suitable safeguards, in order that the general public, clinicians and innovators can believe in how AI is adopted throughout the system.
“And, whereas chatbots like ChatGPT will be useful for normal data, they need to not substitute recommendation from a healthcare skilled,” he mentioned.
Penny Kechagioglou, chief scientific data officer (CCIO) at College Hospitals Coventry and Warwickshire NHS Belief and chair of the Digital Well being Networks CCIO Advisory Panel, mentioned that the survey outcomes are “unsurprising” and clinicians must work with sufferers to perceive how digital and AI can assist them.
“The protected adoption of the suitable AI instruments will fulfill affected person curiosity and sense of management, assist clinicians with their workload and re-build belief between sufferers and clinicians,” she mentioned.
The survey additionally discovered that 38% of sufferers surveyed felt uncomfortable mentioning data discovered on-line to their healthcare supplier.
Christopher Lippuner, co-founder and chief government of Semble, mentioned that sufferers are drawn to those instruments “out of comfort or nervousness”.
“The query is not whether or not sufferers will use AI, however how we will make that use protected, guaranteeing they perceive the bounds and really feel supported to convey any insights into knowledgeable conversations with clinicians,” he mentioned.
Rachel Energy, chief government on the Sufferers Affiliation, instructed Digital Well being Information that sufferers are turning to AI for “well timed data in a system underneath actual stress” and that the NHS should “meet them the place they’re”.
“It’s essential that sufferers aren’t dismissed for doing their very own analysis however are recognised as equal companions of their care. On the identical time, there are actual considerations concerning the security and reliability of well being data on-line.
“To handle this, we have to improve well being literacy, strengthen the regulation of false or deceptive content material, and assist the trusted relationships with well being professionals which are on the coronary heart of excellent care,” Energy mentioned.
