A £330 million contract to wire the NHS's patient data through Palantir's Federated Data Platform is fracturing on contact with reality. Freedom of information responses from more than two dozen NHS trusts and health boards reveal a pattern of rejection, indifference, and reversion to legacy systems — a damning signal for a programme pitched as transformational. From Gateshead to Truro, organisations caring for millions of patients are declining to use the platform or have trialled it and walked away. The language in the FOI responses is unusually blunt for the NHS. Greater Manchester's care board called the system 'duplicative' and 'retrograde.' A Suffolk trust said existing systems 'supercede' the FDP's functionality. University Hospitals Birmingham, serving over 2 million people, cited 'limited expected benefit over existing solutions.' Cambridgeshire and Peterborough said the system 'has not offered significant improvements on local systems.' These are not ideological objections — they are operational verdicts from clinicians and data teams who tested the product and found it wanting. Palantir disputes this characterisation aggressively. The company says 137 trusts have reported benefits, points to 117,000 additional operations and a 14.3% reduction in discharge delays for long-stay patients among FDP users, and notes the programme holds a green delivery rating — placing it in the top 20% of major government programmes. But NHS England itself has added a caveat to those statistics: other variables have not been controlled for, making cause-and-effect claims unreliable. The gap between Palantir's headline numbers and the statistical rigour behind them is where the real argument lives. The political architecture around the deal is intensifying. Two parliamentary select committees have urged the government to end the contract. The national data guardian, Dr Nicola Byrne, urged ministers to weigh the 'trust factor' — more than 44,000 people have filed legal objections to their health data being processed through the FDP. Meanwhile, Palantir appointed former Labour MP Tom Watson as its chief lobbyist, warning his former colleagues not to let 'mob rule lead procurement practices.' The revolving door between Westminster and Palantir's boardroom is now a visible pressure vector on a Labour government already under scrutiny. The core question is not whether Palantir's technology works in isolation — some trusts clearly find it useful for waiting-list management and surgery scheduling. The question is whether a single US vendor should own the integration layer for England's entire health data infrastructure, especially when a significant fraction of end users reject the product. The Medact campaign and competitors like Voror argue that interoperable, home-grown systems would deliver more value without the political and privacy overhead. Andy Burnham's government now faces a binary decision at the break clause. Extending the contract locks in Palantir as a structural dependency for years. Exercising the break risks transition costs and the loss of whatever genuine gains the platform has delivered. Neither option is clean, and the FOI data makes the status quo harder to defend — not because Palantir is uniquely bad, but because the evidence for its unique indispensability is thin. What the FOIs really expose is a procurement model that imposed a centralised solution on a system that resists centralisation by design. NHS trusts are operationally autonomous for good reason; their data environments, clinical workflows, and legacy systems vary enormously. A platform that works brilliantly at Chelsea and Westminster may be useless in Humber. The deeper failure is strategic: betting £330 million on a single vendor for a problem that is fundamentally about interoperability standards, not about any one company's software.
