In June 2024, porters in south-east London began carrying specimens between buildings by hand, and clinicians wrote requests on paper forms that no system could reconcile. A ransomware attack on a pathology provider had cut the link between samples and results. Operations were postponed and blood supplies ran short. Patient data later surfaced online. Few people outside the sector had heard of Synnovis before that week, and few inside it have forgotten the name since.

Against that background, the announcement that Palo Alto Networks will supply cloud-based security through NHS England’s Secure Boundary programme, protecting around two million computers, deserves more attention than a corporate press release usually earns. Markets noted it briefly. Its relevance to the health service is structural.

The NHS has long defended itself as hundreds of separate organisations, each with its own network, patching routine and appetite for risk. A trust with a stretched digital team and a decades-old estate must choose between replacing a failing ward system and upgrading a firewall, and the ward system usually wins. WannaCry in 2017 exposed the consequence, with some 19,000 appointments cancelled across England. Secure Boundary accepts that the problem cannot be solved trust by trust. A national service that inspects traffic and blocks threats before they reach local devices removes a category of work from organisations that never had the capacity to do it well.

The timing matters because government ambition for NHS data has grown considerably. The ten year health plan envisages a single patient record and a more capable NHS App, and it expects wider use of the federated data platform and of AI in clinical and administrative work. Each depends on the public believing that records are safe. Patients asked to accept broader data sharing will weigh the request against the most recent headline they have read. A national security layer gives ministers something concrete to point to. It also gives chief information officers a baseline they can rely on rather than renegotiate every year.

The limits deserve plain statement. A boundary protects the devices inside it. The Synnovis attack reached the NHS through a supplier, and much of the service’s exposure sits with laboratories and software vendors that connect to trusts without belonging to them. Two million managed computers is a large number, yet it leaves a considerable surface untouched. Concentration carries its own hazard. Routing a significant share of the estate through one vendor’s cloud brings efficiency and a single point of dependency, so procurement teams will need to be sure that exit terms, resilience testing and performance reporting are as rigorous as the technical specification.

There is an organisational risk as well. Centralised security can encourage local boards to treat cyber resilience as someone else’s responsibility. Trust leaders remain accountable for clinical continuity, and recent incidents show that recovery depends on rehearsed manual fallback, backups that have actually been restored in testing, and early agreement on which services restart first. None of that can be bought from a supplier. Financial pressure sharpens the temptation, since a national programme can look like a reason to trim local spend in a year when every budget line is under challenge.

Policymakers face a related task. The cyber security and resilience legislation the government has advanced is meant to place formal duties on more suppliers, and its value to health will depend on whether regulators have the staff and powers to enforce it. For health-tech and life sciences companies the direction is clear. Security assurance is becoming a condition of trading with the NHS, and smaller firms without it will find procurement routes narrowing.

Secure Boundary deserves credit as a sensible exercise of state capacity, delivered at national level where the weakness has always lain. Its success will be judged on the day a supplier fails and a trust must keep treating patients. That day will test the programme, and it will test the local plans standing beside it.

‍