Earlier this year, I had the privilege of joining colleagues from across health and technology at 10 Downing Street for a roundtable on the future of AI in Primary Care. Sitting around the table with clinicians, policymakers and fellow innovators, it was clear that the appetite for change is real, but so too are the challenges.
The government’s new 10-Year Health Plan sets out three pillars it hopes will reshape the NHS: Analogue to Digital, Hospitals to Communities, and Sickness to Prevention. On paper these are the right priorities. The question is how to turn words into action.
AI is rightly generating huge attention. Whether it is automated clinical documentation, triage, or patient communication, the promise to relieve pressure on frontline staff is exciting. But not all AI delivers, and hype alone will not help a single patient. Some of the tools making the biggest impact in practices today are not flashy at all. Messaging platforms, call and recall automation, and workflow support are quietly reducing workload every day. We should be celebrating what works and investing in the technologies that have proven value rather than chasing headlines.
The structure of Primary Care technology itself is another barrier. Too much is still dominated by legacy systems that lock practices into rigid ways of working. A modular, interoperable setup would allow specialist vendors to compete and deliver better results. The NHS has an important role to play in building the backbone infrastructure, such as unified patient records and shared standards, but practices should be free to choose the right tools for their needs. Flexibility and competition are far more powerful drivers of improvement than centrally imposed one-size-fits-all frameworks.
Adoption is where the real work begins. Practices that are enthusiastic early adopters can show what is possible and inspire others to follow. Those who are slower to change will still need support, but centrally driven rollouts often struggle to stick. The incentives are not always there. Aligning operational targets and financial flows more directly with patient outcomes and efficiency would change that. When practices see that innovation leads to tangible benefits for their patients and their teams, adoption accelerates naturally.
Regulation is another part of the puzzle. Safety checks are essential, but the way they are handled today is often slow, duplicated and inconsistent. The idea of a regulatory passport is attractive if it is done properly. It could remove duplication, speed up safe adoption and give innovators confidence to invest, while still keeping patients protected. The key will be to keep it simple and consistent, not add another layer of process for already overstretched teams.
The Health Plan sets the right ambitions, but transformation will depend on delivery. Success will come from building infrastructure that allows innovation to flourish, setting clear standards for interoperability, and making sure incentives line up with outcomes. Most of all it will depend on trust. Trust in technology that proves its worth, and trust in the clinicians and patients who use it. If we can create that environment then Primary Care will not just survive the next decade, it will thrive.