Great Ormond Street Hospital has released the results of what it calls a “landmark study” into AI-scribing, and the numbers are striking. Using TORTUS’s AI transcription tool across more than 17,000 patient encounters in hospitals, GP practices, mental health services, and ambulance teams, clinicians were able to spend nearly a quarter more time directly with patients. Appointment lengths fell by more than eight percent, and a third fewer clinicians reported feeling overwhelmed by note taking. In A&E departments, the results were even sharper, with a 13.4 percent increase in the number of patients seen per shift.

Patient attitudes appear broadly supportive. Ninety-two percent of patients consented to AI scribing, with many reporting that consultations felt more engaging. By shifting the focus back to eye contact and dialogue rather than typing, the technology seems to support not only efficiency but also the human connection at the heart of clinical care.

The economic case is equally compelling. Independent modelling by the York Health Economics Consortium found that if each clinician could see just one more patient per shift, the NHS would gain almost 9,300 extra A&E consultations every day. That translates to £176 million in documentation time saved and £658 million in additional clinical capacity annually. For a system under intense financial and operational pressure, these are not marginal gains but potential lifelines.

The response has been swift. Great Ormond Street plans a wider rollout across outpatient settings this autumn, while Health Minister Stephen Kinnock hailed the pilot as “exactly the kind of innovation we need” to tackle backlogs and build a digitally enabled NHS. He emphasised that freeing clinicians from admin is not only about efficiency but also about restoring the quality of patient relationships.

Yet for all the optimism, there are challenges. Scaling any technology across the NHS requires not just investment but trust, training, and robust governance. Questions remain about how AI scribes handle sensitive data, how outputs are validated by clinicians, and whether benefits will be distributed evenly across specialties and regions. There is also a risk of uneven adoption, with well-resourced trusts racing ahead while others struggle to keep pace.

The wider AI scribing market is already heating up. Heidi Health recently raised $16.6 million to expand its own technology into GP practices through a partnership with Modality. Accurx has teamed up with Tandem Health to offer AI transcription and coding directly into EMIS and SystmOne. Meanwhile, independent pilots like the one at Royal Devon are testing alternative providers. This competition may accelerate innovation, but it also highlights the importance of national guidance and standards to prevent fragmentation.

What the GOSH study shows most clearly is that AI scribes are not a futuristic idea but a present reality. They can free clinicians from hours of administrative burden, improve patient flow, and reduce burnout. But if this technology is to deliver on its promise, adoption must be thoughtful, equitable, and underpinned by strong safeguards.

For the NHS, the opportunity is immense. The real test will be whether leaders can turn promising pilots into sustainable, system-wide change without losing sight of the core purpose: giving clinicians more time with their patients.