In February, Louis Mosley, who runs Palantir’s British business, walked into the office of Luke Pollard, the defence procurement minister, carrying a 23-page document. It was a review of how the Ministry of Defence buys software, and its argument was blunt. Cut the paperwork. Relax scrutiny of new deals. Let existing contracts grow without a fresh competition.
Thirty-five recommendations followed. According to the investigative outlet Democracy for Sale, which published the review and the ministry’s internal minutes on Oct. 24, exactly one was rejected outright.
The awkward chronology runs the other way. In December, two months before the report landed, the ministry awarded Palantir a follow-on contract worth £240 million with no competition at all. The company already held a £75 million software deal. Neither the review nor the company’s role in writing it was disclosed at the time. The documents surfaced only after freedom of information requests.
Britain buys a great deal from a small number of firms, and defence procurement is the hardest part of it. The ministry spends billions a year on software that runs logistics, personnel records and battlefield analysis, and its own auditors have complained for a decade that projects arrive late, over budget or both. Reform is genuinely needed. The question the documents raise is who gets to write the reform.
The forum for that conversation was set up in June 2025, when the ministry created a Defence Industrial Joint Council where officials could discuss policy with the bosses of major suppliers. Palantir was one of about a dozen companies invited, alongside three lobbyists, two trade union representatives and the minister. A pledge to include academics produced no seats for universities or civil society groups. The council does not publish its records, which is why the minutes arrived by request rather than by press release.
At the second meeting, in September, the ministry asked Palantir and five other members how it could buy software faster. The answer that came back in February reads like a supplier’s wish list, and the ministry’s own internal reporting subsequently recorded significant progress in adopting it.
Personnel moved along the same track. Barney Kistruck, the ministry’s industrial strategy director, told the first council meeting that procurement reform would be a priority as defence spending rose to what he called record-breaking levels. He left the ministry in August 2025. Nine days later he joined Palantir.
The company’s reach into British institutions is not new, and it is not confined to defence. Palantir, co-founded more than two decades ago by the investor and Trump donor Peter Thiel, holds contracts across analytics and AI-assisted war-gaming. Its £330 million contract with the National Health Service is up for renewal, a decision that sits with Andy Burnham’s government and that has drawn opposition inside the governing party. Last month Tom Watson, a former deputy leader of Labour, joined the company to, in his words, correct false information about it.
Anti-corruption campaigners read the procurement documents as a conflict-of-interest problem rather than a policy one. Sue Hawley of Spotlight on Corruption called for a full National Audit Office review of whether the ministry is following its own rules on conflicts. The audit office has the standing to look at exactly this: whether advice shaped a commercial decision, and whether the advice came from a party with money on the outcome.
Britain’s procurement law gives the ministry room to move quickly when it judges urgency or sole-supplier status, and follow-on awards that expand an existing system are common in software, where switching costs are real. The audit trail is supposed to carry the justification. A £240 million award without competition requires a published notice explaining why no tender was run, and the reasoning is where scrutiny usually stops: a paragraph of official prose that few outsiders read and fewer challenge.
What the minutes add is the sequence. A supplier told the buyer how to loosen its own rules, the buyer adopted most of it, and the same supplier then collected a contract under the looser rules. Each step has a defensible explanation. Taken together they describe a loop that no single document closes.
For smaller British software firms the practical effect is thinner competition in a market where the ministry is one of the largest buyers of anything. For the ministry the risk is narrower and more expensive: rules relaxed on a supplier’s advice tend to be relaxed in that supplier’s favour, and the savings from skipping a tender are easy to book in year one. The costs arrive later, when the system is embedded, the alternatives have gone, and the next contract is negotiated by one buyer facing one seller.
Pollard’s office has not published the review, and the company has not explained why a firm bidding for ministry work was asked to advise on the rules governing its own bids. The documents are now public. The decisions they shaped are not yet.
