UK GPs and generative AI: adoption has outrun governance

Author: Alexander Börve

November 1, 2026

UK GPs are adopting generative AI faster than healthcare organisations can provide the training, guidance and oversight needed to manage its use. For digital primary care services, the challenge is no longer whether AI will be used, but how to introduce it responsibly.

Generative AI is already part of everyday general practice in the UK, but the rules around its use are struggling to keep up. A survey of 1,004 UK GPs conducted in January 2026 found that 52% used generative AI in clinical practice. The study, published in BMJ Digital Health & AI, also found that 70% of those using AI said it had reduced their workload.

The change has been rapid. Earlier rounds of the same annual survey found that 20% of GPs used generative AI in early 2024 and 25% in January 2025. Each survey involved a different group of GPs, so the results are snapshots rather than a record of the same doctors over time. However, they show a clear rise in adoption.

The central concern is that doctors are using AI faster than training, guidance and governance are developing. For digital primary care services, this makes the choice of AI tools important. Each feature needs a clear purpose, defined limits and an accountable person responsible for what happens next.

What GPs use it for

GPs mainly use generative AI for everyday tasks. Among users, 51% used it to write letters, while 41% used it for real-time documentation during consultations. These tools can listen to consultations and prepare draft notes.

Some GPs also use AI for clinical tasks. Among users, around 28% used it to help with differential diagnosis, while 27% used it to explore treatment options.

Many of the tools reported in the survey were not designed specifically for healthcare. The most commonly named tools included a general-purpose chatbot, an ambient scribe and a general-purpose assistant. This suggests that AI is already part of UK general practice, often through tools originally developed for wider use.

Collaborator, tool or competitor?

GPs have different views about AI and its future role. Nearly half of respondents, or 46%, described AI as a helpful collaborator working alongside doctors. Another 21% viewed it as a simple tool, while 10% considered it a potential competitor.

Looking ahead ten years, 34% expected AI to change which skills matter most for GPs. A further 16% expected it to replace some core skills.

Over a 20-year horizon, 59% expected AI to reshape the GP’s role substantially while keeping doctors central to patient care. These findings suggest that many GPs see AI as a way to change how they work, rather than remove the need for doctors.

Governance has not kept pace

The growth in AI use has not been matched by sufficient training and guidance. The clearest figures on employer support and training come from the January 2025 survey. Of its respondents, 85% said their employer had not encouraged them to use generative AI, and 95% reported having no professional training in using it.

A separate survey of 2,108 GPs, published by the Nuffield Trust and the Royal College of General Practitioners in December 2025, found that 28% used AI tools in clinical practice. This included 11% who had obtained the tools themselves. The two surveys used different samples and methods, so their headline figures are not directly comparable. The report found that, outside ambient voice technology, where national NHS guidance exists, AI implementation depends heavily on local policies. Some integrated care boards prohibited all AI use, while others encouraged it.

The authors of the 2026 study concluded that “policy now clearly lags behind practice.” GPs are adopting consumer and partly integrated tools to manage their workload, sometimes before receiving formal approval from their employers. Without suitable training, clear professional boundaries and proper oversight, AI could become a routine part of primary care without being deliberately introduced and managed.

What it means for digital primary care

For digital primary care services, the findings highlight two important points. First, doctors may adopt AI tools independently if approved options do not meet their needs. Services must therefore provide useful tools that clinicians can trust and use confidently.

Second, AI features need clear limits and accountability. Questions about responsibility, oversight and errors apply to both clinical tools and AI features offered directly to patients.

Services should be able to explain what each AI feature does, what it cannot do and who is responsible for deciding the next step. This helps teams choose tools that fit their intended purpose and manage their use appropriately.

Where a dermatology API fits

Autoderm is a dermatology API designed for a specific purpose. Digital primary care services can integrate it into their own apps or websites under their own brands.

Users upload a photograph of a skin concern and receive informational condition suggestions. The service covers a wide range of skin concerns.

Autoderm handles the image analysis and provides suggestions for the person using the app. The digital service remains responsible for deciding what happens next. This gives developers a way to add a defined dermatology function without treating the API as responsible for the entire care pathway.

See how it works as a dermatology API for virtual primary care. Autoderm’s published studies are collected on the clinical evidence page.

Questions for digital primary care teams

Before introducing or expanding AI features, digital primary care teams should ask:

Which AI tools are clinicians already using, and which have been formally approved?
What training and guidance are available for each tool?
Can the team explain what each patient-facing AI feature does, what it cannot do and who decides the next step?
Who is accountable when an AI output is incorrect, and how are errors reported?

These questions can help services identify gaps in their current approach and make more informed decisions about AI adoption.

The takeaway

Generative AI use among UK GPs has grown rapidly, while training and guidance have struggled to keep up. For digital primary care services, the priority is to adopt AI deliberately rather than by default. Every feature should have a clear purpose, defined limits and an accountable owner for the decisions that follow.

References

Blease C, Garcia Sanchez C, Gaab J, Alderman JE, Allen L, Kumpunen S. Generative artificial intelligence adoption and the future of the primary care profession: an online survey of UK general practitioners’ experiences and opinions. BMJ Digital Health & AI, 2026.
Blease CR, Locher C, Gaab J, Hägglund M, Mandl KD. Generative artificial intelligence in primary care: an online survey of UK general practitioners. BMJ Health & Care Informatics, 2024.
Blease C, Hagström J, Sanchez CG, et al. General practitioners’ adoption of generative artificial intelligence in clinical practice in the UK: an updated online survey. Digital Health, 2025.
Kumpunen S, Lobont C, Garrard L, Fisher D, Lau R, Fallica G, Fisher R. How are GPs using AI? Insights from the front line. Nuffield Trust, 3 December 2025.

Alexander Börve

Alexander Börve, MD is the founder and CEO of Autoderm. He has worked in digital dermatology since 2006 and has published seven peer-reviewed papers on mobile teledermatology and AI-assisted dermatology.