The Palantir story every UK supplier should be reading as a warning (and an opening)

There was a headline doing the rounds last week that a lot of people in public sector sales will have skimmed as a health story or a Westminster story. It is neither. It is a procurement story, and it is one of the clearest signals yet of where the whole market is going.

The short version: the Telegraph reported last week that Andy Burnham, widely expected to enter Downing Street later this month, is minded to end Palantir's role in the NHS. At the centre of it is the Federated Data Platform, a seven year contract worth around £330m, awarded in 2023, now used by more than half of NHS trusts in England. There is a break clause next March, which means a decision to walk away would need notice served in December.

A few honest caveats before we build anything on top of it. This is a report, not a decision. Burnham's own allies have been quick to tell journalists the story was not briefed by his camp, and he has made no formal commitment on any existing contract. So treat the specifics as unconfirmed. What is not in doubt is the direction of travel, because it fits a pattern that is now impossible to ignore.

This was never really about one contract

The case against Palantir has been building for months, and it splits neatly into two arguments that people tend to blur together.

One is ethical and political: the company's work with the Israeli military and with US Immigration and Customs Enforcement, the Trump-world associations of Peter Thiel and Alex Karp, and reporting around contractor access to patient data. That is what campaign groups, unions and a chunk of Labour MPs have been pushing on.

The other is structural, and it is the one that matters for the rest of us. Should a US company, subject to US law, hold this much of the plumbing of a national health service? That question does not depend on what you think of Palantir specifically. It is the same question being asked in Paris, Copenhagen, Berlin and Brussels, about a whole category of suppliers.

And it is worth being fair to the other side, because the counter-argument is real. The shadow health secretary has framed any cancellation as choosing "politics over patients." The FDP has been credited with helping deliver something like 110,000 additional operations. Sadiq Khan's decision to block a separate £50m Met Police contract drew rare public criticism from the Met Commissioner, and Palantir has taken it to the High Court. There is a genuine worry, voiced from inside Labour as much as outside it, that a hard turn against US tech spooks one of the few parts of the economy that is actually growing. None of that is trivial. But notice what has happened: provenance has become a live evaluation criterion, sitting right next to cost and clinical outcomes. That is new.

The European backdrop: de-risking from US tech has gone mainstream

If you only watch the UK, this looks like a one-off political scrap. Zoom out and it is the local expression of a continent-wide movement.

In January, France told 2.5 million civil servants they would drop Microsoft Teams, Zoom, Webex and GoTo Meeting in favour of Visio, a homegrown, open source platform, across all of government by 2027. That was the template: one deadline, one sovereign replacement, no opt-outs. It has since extended to a mass migration off Microsoft Windows onto Linux, part of the broader Suite Numérique push to replace US services wholesale.

France is not an outlier. Denmark's digital ministry is moving to Linux and LibreOffice. Schleswig-Holstein in Germany has shifted tens of thousands of workstations off Microsoft. Austria's military has swapped Microsoft Office for LibreOffice. Amsterdam has published a digital autonomy strategy running to 2035. The EU now has a Commissioner explicitly holding a "tech sovereignty" brief, warning that dependence "can be weaponized against us." And France's own domestic intelligence service has already terminated its Palantir contract.

The legal engine underneath all of this is the collision between the US CLOUD Act and GDPR. The CLOUD Act lets US authorities compel American companies to hand over data wherever it is stored. Microsoft's own lawyer told the French Senate last summer that he could not guarantee French data held in European data centres was beyond the reach of the US government. Once that is on the record, no amount of "sovereign cloud" marketing fully closes the gap, because the problem is legal jurisdiction, not the location of the servers. That single admission has done more to move European procurement than a decade of policy papers.

The UK's own sovereign turn

Here is the part that too many commentators miss. The UK is not just reacting to Europe. It is running its own version of the same play, and has been for a year.

The framing is "AI maker, not AI taker," lifted from the January 2025 AI Opportunities Action Plan. Since then we have had a £500m Sovereign AI Fund and Unit, with an £80m pot for government to act as an early customer of British AI, buying prototype capabilities to give startups revenue and reference customers. In June, at London Tech Week, that was topped up with a £1.1bn AI Hardware Plan: a £750m national supercomputer, £400m for advanced chips, £150m earmarked for buying inference hardware from UK vendors this summer. The whole logic is demand-pull. Use the state's chequebook to create a domestic market that would not otherwise exist.

And crucially, the procurement rulebook has been rebuilt to allow exactly this. The Procurement Act 2023, live since February 2025, quietly changed the grammar of how contracts are awarded:

• The old "most economically advantageous tender" test became "most advantageous tender," giving buyers explicit room to weigh things beyond lowest price.
• Social value weighting has crept up towards 20% in many evaluations and is increasingly the deciding factor. Burnham's line about wanting social value to play a bigger role in awards is not rhetoric, it is a lever that already exists in the Act.
• There is a national security exemption that lets departments hand contracts to specific suppliers in sensitive sectors. The government has already leaned on this, pointing to roughly £5bn of contracts flowing to British firms since March.
• New SME and British-industry spend targets, plus a "growing British industry, jobs and skills" consultation, all point the same way.

Put the pieces together and the direction is unmistakable. Where a supplier is based, whose law it answers to, and what social and economic value it leaves behind are moving from "nice to have" to scored, enforceable award criteria.

So what does this actually mean for you

If you sell into the UK or European public sector, or you advise anyone who does, three things follow.

First, sovereignty is now a bid discriminator, not a footnote. Data residency, corporate ownership, jurisdiction and supply-chain provenance belong in your qualification checklist and your written responses. If you cannot speak to them clearly, assume a competitor will.

Second, this is a genuine opening for British and European SMEs, and it is being deliberately engineered. Government-as-early-customer, SME spend targets and the national security exemption are all designed to tilt the field toward domestic suppliers. The catch is that "buy local" only survives contact with reality if the local option actually delivers. The "politics over patients" attack lands precisely when the sovereign choice is worse. So the winning move is not to wave a flag, it is to be demonstrably good and British, in that order.

Third, watch the tension, because it is unresolved. There is a real risk of procurement becoming performatively protectionist, of decisions dressed up as sovereignty that are really just politics, and of the UK talking itself out of world-class tech at the exact moment it says it wants to lead. Mature buyers will hold two ideas at once: reduce strategic dependence on any single foreign supplier, and refuse to sacrifice outcomes to do it.

The Palantir headline will be litigated politically for months, and the specific decision may yet go either way. But the underlying shift is already done. Across the UK and Europe, the question buyers are asking has quietly changed. It used to be "who is cheapest and best?" It is becoming "who is cheapest, best, and ours?" If you are in this market, that third word is the one to plan around.