UK ministers are reportedly considering triggering a break clause in Palantir’s £330 million contract for the NHS Federated Data Platform (FDP). The clause, available from February 2027, would allow the government to exit the seven-year deal early.
The review follows mounting pressure from the British Medical Association, MPs and campaign groups, who have raised concerns about data privacy, conflicts of interest, and Palantir’s links to the US defence and security establishment. Critics have gone further, describing the company as “ethically bankrupt” and highlighting its involvement in US immigration enforcement.
What the FDP Actually Does
The NHS Federated Data Platform is, in principle, a fairly sensible idea.
It brings together data from across the health service into a single system, allowing staff to make better decisions about patient care and service planning. At the moment, NHS data is fragmented across multiple incompatible systems. The FDP is designed to fix that.
It operates at three levels:
- National (NHS England): combining data from legacy systems to support national planning
- Regional (Integrated Care Boards): helping local systems understand population needs and allocate resources
- Local (NHS Trusts): giving frontline staff access to up-to-date patient information, improving scheduling, waiting list management and care planning
None of this is especially controversial. If anything, it is long overdue.
Why Palantir?
The FDP grew out of the Covid pandemic, when the NHS needed to rapidly integrate data across a highly fragmented system. Palantir initially worked for free to build the Covid Data Store, before being awarded a £23.5 million contract to continue that work.
In November 2023, a Palantir-led consortium won the £330 million FDP contract, based on its ability to meet NHS England’s technical requirements. The platform is built on Palantir’s Foundry software, which specialises in connecting incompatible databases and analysing large, complex datasets.
The contract covers up to 240 NHS organisations, including trusts and integrated care systems.
Importantly, the agreement places strict limits on what Palantir can do with NHS data. It cannot:
- commercialise or market NHS data, even in anonymised form
- use NHS data to develop new products
- train AI models on NHS data
Palantir retains intellectual property rights over its core software, but the NHS has built protections into the contract to prevent misuse of patient data.
Why the Backlash?
The controversy isn’t really about the technology. It’s about the company.
Palantir has long attracted criticism for its work with intelligence agencies, defence organisations and law enforcement. Its culture is unapologetically aligned with the US security state.
That history matters.
In 2013–14, a Palantir employee worked with Cambridge Analytica on tools that harvested data from millions of Facebook users. More recently, the company has been an outspoken supporter of Israel during the Gaza war, signing agreements to support “war-related missions”. Its chief operating officer has talked openly about a future in which Palantir software is embedded “inside every missile, inside every drone”.
Last week Palantir published a 22 point manifesto, denouncing inclusive societies, and positioning itself as the defender of the West. This looks like the country aligning itself with America’s new foreign policy model which rejects multiculturalism, tolerance and liberal values.
None of this sits comfortably alongside a publicly funded health service. The NHS has it’s own values, which are completely at odds with Palantir. That doesn’t mean they can’t work together, but it does ask the question why anyone thought that they were a good strategic fit for the NHS in the first place.
But there is a complication with all of this. There is no suggestion that Plantir is failing to deliver it’s contract or has doing anything bad with our data. It is purely a reputational decision.
A Wider Shift in Procurement
The review of the Palantir contract also needs to be seen in a broader context.
The UK’s new procurement regime, under the Procurement Act 2023, came into force in February 2025. It is designed, at least in part, to make it easier for smaller firms to win public contracts — but also to give government more flexibility to bring services back in-house.
The incoming Labour government has leaned into this. There is a clear emphasis on “resilience”, supply chain security, and reducing dependence on large external contractors.
In parallel, the reintroduction of the two-tier workforce code aims to ensure that outsourced public services do not create worse conditions for private sector hires compared to transferred public sector staff.
But it also means that Government Departments can get rid of suppliers and subcontractors based on reputation. This was always a frustration for me when I ran Government commercials. A firm could fail and fail again, embarrass the Government over and over again and yet still get more business. Each case had to be treated on it’s merits, which meant that the same firms got the same business and failed in an endless cycle. A firm which failed for MOJ and DH couldn’t be stopped from being awarded contracts by DWP.
The new legislation changes this, and Plantir is the first big test. They are being dropped as a supplier, not on cost or performance but because of their reputation, and the risk that they will embarrass HMG.
Taken together, this points to a subtle but important shift: away from outsourcing as the default, and towards greater control within the public sector. Last year Rachel Reeves flexed the Governments debt target allowing higher levels of capital spending. All of these things are boring technical changes that only people fixated on public finances would notice. But they are part of a gradual re-writing of the rules of the state so that cuts to public infrastructure and a default preference for outsourcing are no longer baked in.
I have no doubt that the vicious attacks on Reeves, and to an extent Starmer are funded by big companies who make a lot of money out of taking over the state, and who see the current Labour Government as the biggest global challenge to their business model.
So What Happens Next?
The break clause in 2027 gives ministers an option, not a decision.
The FDP itself addresses a real problem: the NHS’s fragmented data infrastructure. Walking away from it would not make that problem disappear. If anything, it would raise an awkward question — what replaces it? Who takes on the contract?
At the same time, the political pressure is unlikely to go away. Palantir is not just another IT supplier. It carries baggage, and that baggage matters in a system built on public trust.
The government is therefore caught between two fairly unappealing options:
- stick with a technically capable but politically controversial supplier
- find another supplier with all the cost and disruption that implies
Neither is particularly attractive.
Which probably explains why, for now, this is a “review” rather than a decision. But this shouldn’t detract from the sea change – a Government who are prepared to face down big outsourcers, and remove contracts from suppliers who don’t perform, or whose operations carry reputational risk.
When I ran Government procurement we knew that certain suppliers had a terrible reputation, and yet they were able to work the procurement rules so that they kept winning contracts. Those days are over.
But if the NHS sack Palantir for it’s far right ethos and it’s links to far right Governments, that opens the door for a future Government to take contracts away from suppliers and giving them to their chums and donors for being too woke or not “British” enough. Although the current rules didn’t stop a previous Government from giving billions of contracts to their chums and donors during Covid did it?
https://www.stoneking.co.uk/employment-rights/public-sector-outsourcing