Two letters to the Guardian, written independently from Cheltenham and Shropshire, converge on the same structural diagnosis: the UK has no digital sovereignty, and the problem runs deeper than any single contract. George Williamson at the Alan Turing Institute warned that foreign AI technology can be switched off at short notice. The letters extend that logic to operating systems, cloud services, and the entire NHS data stack. The threat model is not theoretical. The US government imposed targeted sanctions on the International Criminal Court, cutting digital services to specific individuals. Ian White's letter makes the jump that most commentary avoids: if this can happen to an international institution, it can happen to a country. Every software update your machine downloads is a dependency. Every unpatched bug is a door. Holland and France have already begun building nationwide IT ecosystems they control, actively breaking American tech's near-monopoly. The UK has not. White frames this as a straightforward sovereignty gap — not protectionism, not techno-nationalism, but the minimum viable condition for a state that wants to make independent decisions. Bill Cawley's letter from Shropshire attacks the same problem from the NHS angle. The £330m Palantir contract is the visible symptom, but the disease is a procurement reflex: whenever government needs joined-up data, it buys one large platform from one supplier. These projects consistently overrun on cost and timeline, and they lock the buyer into a single vendor's architecture. Cawley's proposed alternative is elegant and proven: the World Wide Web model. Define an open specification for exchanging records. Build a registry of where each patient's data is held. Let any authorised clinician assemble the full picture, while hospitals and practices choose their own software. Specialist analytics firms compete on the data — but none owns the plumbing. The NHS already has interoperability standards that point in this direction. The question both letters circle is why political attention and money flow to proprietary platforms rather than to making open standards mandatory. The answer is legible through a GRIN lens: large single-vendor contracts are easier to announce, easier to manage politically, and generate concentrated lobbying pressure from incumbents. Open standards generate diffuse benefits for many actors — exactly the kind of value that procurement systems are bad at rewarding. If the pattern continues, the UK's digital infrastructure becomes a series of chokepoints controlled by entities whose interests may diverge from British interests at exactly the moment it matters most. The 20-year trajectory is not gradual decline — it is normalcy punctuated by a crisis in which the off-switch is held by someone else.
