A survey published this month found that 90% of NHS clinicians now report using AI in clinical practice — a striking number given how recent and cautious most institutional NHS AI rollouts have been. It lands alongside confirmation of how the NHS’s £10bn technology funding over the next three years will be spent, money the government says will deliver roughly half the commitments in its 10 Year Health Plan and generate £41bn in benefits over the following decade. A new AI triage tool is also rolling out through the NHS App, reaching over 200,000 patients within a year and every NHS App user by April 2028.
The gap between “90% of clinicians already use AI” and “the NHS is only just funding and formally rolling out AI systems” is the real story. It suggests most current clinical AI use is informal — individual clinicians reaching for general-purpose tools on their own initiative — running well ahead of the sanctioned, procured, audited systems that trust boards and regulators are only now building the funding and governance structures for. That gap is exactly where risk concentrates: unmanaged AI use inside clinical workflows, with none of the safeguards a procured system would carry.
So what
For healthcare organisations, this cuts two ways. It confirms clinician appetite for AI tools is not the barrier — adoption is already happening, informally, whether or not there’s a sanctioned system to channel it into. And it makes the case for moving deliberately: a narrow, well-scoped, properly governed tool — the kind recent NHS trust deployments have favoured over broad chatbot rollouts — gives clinicians a sanctioned alternative to informal tool use, with an audit trail and clinical safety case behind it. If your organisation is weighing where to start, that’s the conversation worth having before procurement decisions get made for you. More on how we approach this at healthcare software development, or get in touch to talk through a project.