At University Hospital Southampton, 185 clinicians and administrators now hold Copilot licences, a market engagement for outpatient administration automation has already concluded, and the trust reports a procurement lead time of three to six months before a supplier can be brought on. Down the road, Humber Teaching NHS Foundation Trust has drawn up its own digital roadmap for the year ahead, promising a four to six per cent cut in administrative time through AI-supported documentation. Meanwhile a five-day accelerator has launched in Bristol to walk digital health innovators through the particular demands of general practice. Three separate organisations, three separate processes, three separate sets of criteria for judging what counts as a success worth repeating elsewhere.

This week Ireland's Health Service Executive published something England does not currently have: a single document setting out how the whole system should identify, fund, test and scale an innovation, with named governance arrangements reporting quarterly to HSE's senior leadership, a stage-gate process with defined go or no-go criteria, and a resource allocation model attached to each phase. It is not a glamorous piece of policy. It reads like what it is, a plumbing document for a health system that has decided duplication of effort is itself a cost worth eliminating. But its existence says something uncomfortable about the system across the Irish Sea that has, for the past eighteen months, been engaged in the opposite exercise.

NHS England's abolition is proceeding through the Modernisation Bill now before Parliament, with functions due to transfer into the Department of Health and Social Care and headcount across the two organisations cut by roughly half. Buried in that legislation is a duty on the Secretary of State to promote innovation, backed by provision for payments and prizes. It is a statutory instruction with no visible mechanism behind it. Sir Jim Mackey's transformation team has its hands full absorbing NHS England's functions rather than building new ones, and the Health Innovation Network that succeeded the old Academic Health Science Networks remains a regional patchwork rather than a national evaluator with the authority to say which pilots deserve national funding and which should quietly end.

The result is visible in the pattern above. Southampton's automation programme, Humber's roadmap and Bristol's accelerator are not evidence of a system failing to innovate. They are evidence of a system innovating constantly, at pace, without anywhere for the lessons to land. A trust that proves ambient voice technology reduces documentation time by a given margin has no obligation, and often no practical route, to share that evidence with a trust three counties over embarking on the identical pilot. Procurement timelines get repeated rather than shortened. Evaluation templates get reinvented rather than adopted. The capital and workforce time this consumes is not trivial in a system already being asked to close a productivity gap against a backdrop of falling central staff numbers.

Ireland's HSE is a smaller system with its own long list of operational failures, and nobody serious would hold it up as a model of healthcare delivery. That is precisely what makes this comparison worth making. The framework it has produced required no reorganisation, no primary legislation and no cut to central capacity. It required someone with the authority to write down a common process and the standing to make it stick. England's Health Bill grants a duty to promote innovation without granting anyone the equivalent standing to build the scaffolding that duty needs. Yvette Cooper inherits a Bill built around consolidating data and commissioning power in the Secretary of State, but consolidation of authority is not the same exercise as construction of process, and the difference will show up first in exactly the kind of trust-level duplication now on public display.

A statutory duty to promote innovation is not itself an innovation strategy. It is an instruction waiting for an architecture. Dublin has just built one without being asked to reorganise anything at all.