NHS AI scribe rollout should include national oversight, report finds.

• Participants in a recent public dialogue said there should be appropriate national oversight of AI scribes as they are rapidly rolled out across the NHS.
• Current low awareness of AI scribes caused frustration amongst participants which must be addressed.
• Participants expected staff to be appropriately trained to use AI scribes.
• Participants thought AI scribe performance should be monitored for accuracy and performance, while recognising that no AI Scribe will be perfect.

AI scribes are AI tools designed to listen to healthcare consultations and generate clinical notes, summaries, and documentation to support doctors with administrative tasks. The dialogue was commissioned in response to growing national policy support for deploying AI scribes across the NHS, despite limited public discussion about expectations from patients on how the technology is used in practice. 

Hospitals are increasingly adopting AI to record consultations and generate clinical notes, often highly sensitive in nature, making it important also to understand how the use of AI scribes in healthcare can be made trustworthy and acceptably safe. If AI is to play a larger role in healthcare, it is critical these early AI deployments are handled carefully: future public confidence in more advanced forms of AI will undoubtedly be shaped by how successfully AI scribes are introduced today.

“I feel that AI can be a huge step forward in the NHS. It would be vital that any conversation between patient and NHS Professional is very accurate and AI must reflect that.”
- Participant 

Previous reports have shown that the public held established views about AI and the NHS, however there was low awareness of existing use of AI scribes. Many participants thought it was important that patients were told the technology was being used and given the option to opt out, aligning with recent guidance from NHS England. 

Participants also highlighted the need for clinician discretion on when it would be more appropriate to not use AI scribes, for example while discussing sensitive topics such as domestic abuse. Greater public awareness of AI scribe use was recommended. 

Many participants were concerned about the quality and accuracy of the data, and that mechanisms such as spot checks or audits were necessary for accuracy. They understood that the accuracy of scribes would unlikely be perfect, but expected the NHS to be ultimately responsible for the level of accuracy, not the AI scribe vendor. 

“...you expect there to be an independent body. That could be repurposing one or charging one that already exists with that purpose.”
‍- Participant 

National governance of AI scribes was expected to ensure good practice is followed. This would include checks on data security, staff training, and accuracy of AI outputs. They wanted national bodies to provide assurance, identify and escalate risks, and support improvement of the technology. 

"AI scribes show promise for the NHS, but a significant voice has been absent from the conversation about the implementation of AI scribes: the patient. This work helps to fill this gap. I extend my thanks to the participants from across the UK for sharing their time and lived experiences, and I hope that this report will help to provide a roadmap to ensure transparency, equality, and safety within this space."
- Professor the Lord Darzi of Denham OM KBE PC FRS, Director, Institute of Global Health Innovation, Imperial College London

“These findings reinforce a simple truth: in the ever-evolving AI landscape healthcare must continue to work for patients first. Transparency, meaningful choice and responsible handling of data are key to protecting patient safety, reducing inequalities and building confidence in the future of care. Above all, patients’ experience must remain central with technology supporting, rather than disrupting, the relationship between patient and clinician. Embedding these expectations is essential to delivering safe care and maintaining trust.” 
‍- Prof Henrietta Hughes, Patient Safety Commissioner for England and Deputy Chair of the National Commission into the Regulation of AI in Healthcare

"The patient and public’s response feels familiar and reasonable. They want to be aware this is happening and they want proper oversight. They don’t expect things to be perfect, but to be good enough. Don’t you feel that way, too? "
- Robin Carpenter, Head of AI Policy and Governance, Newton’s Tree

The public are asking important, practical questions about AI in healthcare, about what it’s for and how it will affect their care. 

These findings matter for those working to enable the adoption of AI in the NHS. They highlight what will be needed to make implementation safe, practical and trusted: strong safeguards, proper training, clear accountability, and a commitment to equitable access. 

"For the AI Commission, this provides a clear steer that has been central to our principles throughout considering a future regulatory approach to AI in healthcare: public trust will depend not just on whether AI works, but on how it is introduced, communicated and governed. Getting that right will be critical to ensuring these technologies deliver real value for patients and the system.  This is about continuing to keep patient perspectives at the centre of decisions that impact their care."
- Prof Alastair Denniston, Professor of Regulatory Science and Innovation, Executive Director of CERSI-AI and Chair of the National Commission into the Regulation of AI in Healthcare

‍

The public dialogue was delivered independently by Hopkins Van Mil (HVM) and supported via a grant from CERSI-AI. Read the full report here.

‍

Media enquiries:           

Lucy O’Neill
lucy.oneill@newtonstree.com

About Newton’s Tree:          

Newton’s Tree is a global healthcare AI company that enables healthcare providers to select, test, deploy and monitor in-house and third-party AI products through its enterprise AI platform. They believe that the sustainable delivery of healthcare relies on radically reimagining healthcare delivery through the large-scale adoption of safe and effective technologies. They work with the leading health systems across the globe to develop and deploy the very best technology.  

About the dialogue: 

The public dialogue was commissioned by Newton’s Tree and delivered independently by Hopkins Van Mil and support from CERSI-AI. The dialogue explored public attitudes towards the use of AI Scribe tools in healthcare, with the aims of this deliberative process to engage and work with a diverse range of participants to discuss the public's views on how AI scribes should be used in primary and secondary care. The dialogue was overseen by a board of representatives from academia, charity, independent research, and not for profit organisations, as well as patient representation. The Board was Chaired by a retired member of NHS staff.

About HVM

Hopkins Van Mil is a social research agency specialising in deliberative processes which bring people together to explore and understand society’s challenges. For over twenty years we have designed and facilitated public dialogues including people in open and constructive conversations to build mutual respect and understanding. Our work which includes people across society, leads to actionable insights, collaborative solutions and evidence-based policy making. 

About CERSI-AI

CERSI-AI is the UK’s Centre of Excellence for Regulatory Science and Innovation in AI and Digital Health. Since launching in February 2025, the Centre has established a national capability for regulatory science in AI and digital health, delivering research, tools, education and innovation support to enable the safe, effective and scalable adoption of AI in healthcare. Through collaboration with the NHS, regulators, academia, industry and patients, CERSI-AI supports AI readiness, regulatory innovation and emerging challenges in AI safety and governance, helping to strengthen the UK’s leadership in trusted AI for health and care.