The new RCP view on digital and AI report exposes a disconnect between doctors’ demand for AI and institutional capability and raises serious questions about the NHS's readiness to adopt AI tools safely. It recommends that the government and NHS must optimise digital systems, ensure digital interoperability, and provide robust regulatory frameworks to protect patient safety, while ensuring doctors and patients can safely benefit from the potential of AI.
-
The RCP view on digital and AI report reveals that while a majority of doctors support widespread implementation of AI in the NHS, 68% believe the NHS lacks the digital infrastructure to introduce it in a way that will make a difference for patients.
-
70% said the NHS’s inability to integrate AI tools with other existing digital systems such as the Electronic Patient Record was the leading barrier to rolling out AI in the health service. Over a third (36%) said a lack of regulation was a barrier.
-
The survey also found 79% of doctors said they needed training in clinical AI tools, yet 66% had no access to such support.
-
69% of 305 UK physicians said they use personal access to AI tools to ask clinical questions.
-
The RCP says the findings highlight a mismatch between physician demands for AI tools and the health service’s readiness to adopt them.
-
The college is calling on the government and the NHS to improve and optimise digital systems to get them AI-ready, alongside robust regulation and giving clinicians time to engage in developing AI tools, so they solve real world clinical problems.
Seven in ten physicians are supportive of AI tools being widely implemented in the NHS, but a similar proportion say it lacks the digital infrastructure to introduce AI that will make a positive difference for patients, new survey data from the Royal College of Physicians (RCP) reveals.
Published today (15 January) in the RCP view on digital and AI report, a snapshot survey of RCP members found 70% were either very (29%) or somewhat (41%) supportive of AI tools being widely implemented in the NHS, but 68% said it lacks the digital infrastructure to introduce AI.48% strongly disagree that the NHS is ready to integrate the tech.
70% of the physicians surveyed by the RCP reported the NHS’s inability to integrate AI tools with other systems such as the electronic patient record (EPR), and that this is the leading barrier to rolling it out in the health service (of 541 respondents).
There is currently no standard across the NHS for electronic patient records, meaning test results are displayed differently across organisations. This places significant and unnecessary administrative burden on clinicians, prolongs treatment and increases the risk of error. The RCP is calling for the NHS to set an EPR content model to improve standardisation and for AI tools to integrate with EPRs.
The RCP says getting the digital basics right, prioritising usability and interoperability between systems, is essential to creating the digital foundations needed for AI to be successful.
The survey also revealed 4 in 5 doctors (79%) said they need training in clinical AI tools, yet 2 in 3 (66%) said they had no access to such support. Alongside the planned changes to curricula1, clinicians must be offered ongoing training to build AI confidence and build the digital clinical leaders of the future.
The RCP report also calls for NHS guidance on the safe use of AI tools in healthcare, with its survey finding that 69% of 305 UK physicians say they use personal access to AI tools like ChatGPT and Microsoft Copilot for clinical questions. The college says the NHS is not moving quickly enough to provide clinicians with safe, approved AI tools. It highlights the urgent need for guidance, education, regulation and more agile NHS procurement processes to provide doctors with approved, effective technologies.
The RCP said the survey exposes a disconnect between doctors’ demand for AI and institutional capability and raises serious questions about the NHS's readiness to adopt AI tools safely. It recommends that the government and NHS must optimise digital systems, ensure digital interoperability – the ability for different digital systems to interact – and provide robust regulatory frameworks to protect patient safety, while ensuring doctors and patients can safely benefit from the potential of AI.
Dr Anne Kinderlerer, RCP digital health clinical lead who led the report, said:
‘Physicians think the NHS is fundamentally unprepared for AI because its digital foundations are broken. Many systems can’t talk to each other, infrastructure is outdated and there is poor standardisation in the function of electronic patient records. This creates huge inefficiencies, frustrations, adds administrative burden and cognitive load on an already over-stretched workforce, delays patient care and increases clinical risk.
‘AI has significant potential but it is not a panacea, and we must guard against optimism bias. We must avoid simply chasing emerging innovative technologies like AI at the expense of optimising existing digital systems and considering how they will integrate with new technologies. Having digital systems that work and link seamlessly, alongside clinician and patient involvement, is critical, so that AI tools can be integrated and deliver on their promise to improve safety and efficiency.
‘Addressing these foundational digital challenges first is critical to ensure that any investment in AI delivers meaningful improvements in patient outcomes and clinicians' working lives. Just having AI tools will not deliver change; that will require effective implementation on strong foundations of interoperable digital systems and complete datasets.’
In the report published today, the RCP has called for:
-
Government to invest in well-functioning digital infrastructure to bring archaic IT systems up to date
-
The NHS to set an EPR content model specification standard that EPR providers must meet to ensure that their products meet NHS requirements.
-
The Department for Health and Social Care to establish a central bank of NHS approved algorithms, AI tools and patient facing apps that meet national standards.
-
Engagement with clinicians and patients from the outset to ensure that digital and AI solutions address real world clinical challenges that improve workflows and support safe patient-centred care.
-
The government’s ‘Roadmap for AI in the NHS’ to set out an effective and ethical plan for how and why AI will be used in the NHS, including how it can enable clinical research and tackle health inequalities, improve interoperability and incentivise the co-design, development and deployment of AI tools with clinicians and patients.
-
Robust regulatory frameworks that put safeguards in place to ensure the safe and ethical use of AI.
Professor Mumtaz Patel, RCP president, said:
‘At a time when physicians are already overstretched, properly integrated AI could revolutionise their working life. The current outdated systems slow clinicians down, forcing them to work round a system that is not fit for purpose. When even basic NHS systems are not working, it will breed scepticism of the potential of AI has to improve care for patients.
‘AI will be most beneficial when it’s designed to support, not replace, clinical judgment. We must train physicians to be the digital and AI leaders of the future and ensure they are involved in the development of AI tools that can help to solve real-world healthcare problems effectively.
‘Right now, it’s a free-for-all. We need AI regulation that protects doctors and patients, along with NHS approved AI tools, apps and algorithms. Having a modern digital infrastructure with effective frameworks in place is essential to unlocking improvements in outcomes and for efficiency right across the NHS.’
Dr Ash Bassi, RCP regional adviser and a consultant physician in St Helens, said:
‘AI is already beginning to change our clinical practice. In gastroenterology, it is being used during colonoscopy to improve the detection of polyps and early cancers, directly enhancing patient safety. For patients, the benefits are clear – earlier diagnosis, fewer missed lesions and more consistent standards of care.’
Professor Jack Kastelik, RCP regional adviser and a consultant physician in Hull, said:
‘Currently, the use of AI is mainly in radiology and imaging. It can also support clinical research. If introduced carefully, it could help to provide doctors with up-to-date research data and clinical evidence to support diagnostics and treatment decisions, as well as supporting administrative workload.’
Dr Damodar Makkuni, RCP regional adviser and a consultant physician in Norfolk, said:
‘AI tools have the potential to support clinicians to answer patient related queries, decide evidence based treatments and make faster, more informed decisions. However, if clinicians become too reliant on AI outputs, this risks errors, bias and it could erode core clinical competencies.
‘At the same time, many proven and safe AI applications – including tools for diagnostics and triage, risk prediction and monitoring, and administrative or ambient documentation – are not yet widely available across the NHS. Expediting access to these tools, alongside strong governance and education, will be critical to realising AI’s benefits safely.’
Dr Alastair Gilmore, RCP regional adviser and a consultant physician in Merseyside, said:
'As doctors, our work depends on rapidly gathering information, documenting care and making diagnoses – processes that are increasingly shaped by decision aids and algorithms. While AI tools promise to streamline this work, adoption across NHS trusts remains cautious. Outdated IT infrastructure and rigid governance often mean new technologies arrive without the assurances or functionality clinicians need to use them safely on the frontline.
'AI-assisted scribing and decision support could, in theory, return valuable time to patient care, but experience has taught us to be careful. Past innovations, including electronic patient records, have often slowed consultations despite their digital benefits. AI must be evidence based and rooted in clear ethical principles. With the right regulation, infrastructure and safeguards, AI can become a tool that clinicians and patients genuinely trust and welcome.'
The RCP’s paper recognises that digital innovations bring risk. It calls for the digitisation of NHS systems and integration of AI tools to be underpinned by robust cybersecurity and governance measures, that safeguard patient information and protect against system-wide disruption.