Stanislas Niox-Chateau, Co-founder and CEO of Doctolib

At eight on a Monday morning, a practice phone line is already queuing. Reception staff work through online requests that arrived overnight, the duty doctor faces a list that has grown since Friday, and somewhere in the building a clinician dictates notes that will be typed up after the last patient leaves. Nothing about this is unusual. English general practice is defined by its routine, and that routine is precisely where a French scheduling company, now backed by the world’s most valuable chipmaker, has decided to operate. 

Doctolib, which counts more than 40,000 GPs among its European users, has received an investment from Nvidia to hire AI specialists and develop models alongside the chipmaker's engineers. It says it has used Nvidia technology for a year on administrative tasks, is building clinical decision support assistants, and plans to extend its patient app into a personal AI health assistant. In May it pledged more than £100 million to Medicus, including 150 new posts and a research centre in London. Together the two moves show a well-capitalised European supplier intending to compete for the software that runs NHS primary care.

That matters because GP IT in England has long been a narrow market. Most practices are served by a small number of clinical systems, and switching between them is slow and costly, a particular burden for teams already short of time.  A new entrant with deep funding and experience in scheduling, documentation and patient follow-up could loosen that arrangement. Many practices would welcome the competition, and so would ministers who want better returns on technology spending. 

The political timing suits the supplier. The government’s ten-year health plan presents digital tools as the means of moving care into communities, with the NHS App as the front door and AI as a way of releasing clinician time. Ambient voice documentation has already spread across practices and trusts, in some places ahead of consistent national assurance.  Doctolib's stated focus on documentation support, workflow automation and administrative help lands where the NHS wants quick productivity gains and where clinicians feel the burden most.

The harder questions sit with the patient-facing ambition.An assistant that can answer any health question and guide people through each stage of care overlaps with what the NHS App is meant to become. If a private platform builds the more useful interface, the state may find that its relationship with the patient, and the data that relationship generates, lies outside its control. Public trust in how health data is handled remains fragile, as earlier data-sharing programmes showed. An assistant that gave poor advice, or quietly favoured particular services, would raise regulatory questions that today’s medical device rules cannot readily answer at scale. 

There is a case for cautious optimism. Doctolib says it is building on open-weight and open-source models and investing in European infrastructure. That could ease worries about reliance on a few American model providers and make independent scrutiny of the models easier. Dependence has not disappeared, though.Nvidia provides the chips and owns a share of the business. Anyone claiming sovereignty should be asked where the data is processed, who can inspect the models, and what happens to the code if the company’s fortunes change. 

For integrated care boards and practice groups, a new supplier is a procurement opportunity that should be used early. Interoperability standards, data portability, clinical safety evidence and liability terms need settling before contracts are signed, because afterwards the bargaining position shifts to the vendor. Practices should also ask what the pilot evidence shows about time saved, since administrative gains are easy to promise and harder to measure across a working week.

Policymakers face a sharper decision.NHS England is being absorbed into the Department of Health and Social Care, which leaves fewer central teams to write standards for a market that is moving quickly.  Without clear rules on assurance and data use, the terms will be set by whichever supplier reaches enough practices first.

Investment on this scale is a vote of confidence in the NHS as a market, and that is flattering. It also hands initiative to the investor. The strongest position the state holds is the one it has today, while it is still being courted, and the moment to define what it requires in return is before the first contract is signed.