A significant survey of UK general practitioners has revealed that one in four are now using artificial intelligence in their clinical work, despite an overwhelming majority receiving no formal training or employer guidance.
Rapid Adoption Without Support
The largest year-on-year survey examining generative AI use among British GPs shows a marked increase in doctors employing tools such as ChatGPT in their daily practice. The research, published on Tuesday 25th November 2025 in the Digital Health science journal, indicates that 25% of GPs now utilise AI technology, up from just 20% the previous year.
Perhaps more concerning is the finding that 95% of these AI-using doctors reported having no professional training in how to effectively and safely deploy these systems. Furthermore, 85% stated that their employers had not encouraged or facilitated the use of artificial intelligence in clinical settings.
How GPs Are Using AI
The survey of 1,005 GPs across the United Kingdom detailed several key applications where artificial intelligence is being deployed:
- 35% of doctors use AI for writing documentation
- 27% employ the technology for differential diagnoses
- 24% utilise AI systems for treatment planning or referrals
Dr Charlotte Blease from Uppsala University in Sweden and Harvard Medical School commented on the rapid shift in attitudes, stating: "In just 12 months, generative AI has gone from taboo to tool in British medicine. Doctors are using these systems because they help - not because anyone told them to."
Significant Risks Identified
Researchers from multiple institutions including Uppsala University, Basel University, the Karolinska Institute, the University of Manchester and Harvard Medical School highlighted several critical concerns about the current unregulated use of AI in healthcare.
The technology's tendency to 'hallucinate' or generate incorrect information presents a clear danger in medical contexts where accuracy is paramount. Additionally, researchers warned about the potential for 'algorithmic discrimination' due to biases in training data that could lead to unequal treatment outcomes.
Patient data privacy emerged as another significant concern, with AI systems potentially handling sensitive medical information without proper safeguards.
Dr Blease emphasised the urgency of the situation: "The real risk isn't that GPs are using AI; it's that they're doing it without training or oversight. This should be a wake-up call. AI is already being used in everyday medicine. The challenge now is to ensure it's deployed safely, ethically, and openly."
Sky News has contacted NHS bodies across the UK for comment on these findings and to ascertain what guidance or training might be developed to address these emerging challenges in primary care.
