29 English hospitals "benefit" from Palantir. They say they don't use it

Plugged in, not switched on

Torbay hospital, in Devon, doesn't hide it: it is "not currently operationally using the FDP", the data platform Palantir built for the NHS. NHS England, the body running the programme, still lists it among connected trusts, and as one realising benefits. Its reasoning, as the Guardian reports it: the infrastructure is there "to support data submission, rather than active operational use of FDP apps."

Torbay isn't alone. According to the Guardian's investigation, published on 2 November, 29 trusts that NHS England presents as benefiting say the system isn't in use at their site, or that they use none of its apps. University Hospitals of North Midlands goes further: "FDP is not live. We do not intend to roll it out in the future."

Think of a gym that counts every badge-holder as an active member, including the ones who never go through the turnstile.

£330 million and a 2027 exit

If you haven't followed this: back in May we covered how Palantir landed at the NHS and at ICE on the same day. The Federated Data Platform is a £330 million contract. It's meant to link scattered patient records, speed up discharges and fill operating theatres more efficiently.

The contract has a break clause in 2027. According to the Guardian, Chi Onwurah, who chairs the Commons science committee, wants the government to use it, and the health committee has also called for Palantir's removal. It's like a lease you can end on a fixed date: the closer it gets, the more every number becomes an argument.

The 29 answers come from more than 100 freedom of information requests filed by Medact, a doctors' group campaigning to get Palantir out, and handed to the Guardian. We haven't read those responses; the count of 29 is the paper's. The official list, though, is public. So we opened it.

We opened the list

NHS England's page lists 142 trusts connected to the platform, the organisations that run one or more hospitals. The ones "realising benefits" are in bold. We counted: 137 are in bold, only five aren't. That's 96% of connected trusts classed as benefiting.

Epsom and St Helier, Mersey Care, Nottingham, North Midlands and Torbay: five of the trusts the Guardian cites are on the list, in bold. A category almost everyone gets into is like a race where every runner goes home with a medal. It no longer tells you who ran.

What "benefits" means is spelled out on the neighbouring page. They come from different kinds of products. Some serve patient care. Others, in NHS England's words, "relate to data analysis and submission tools that support analysts and NHS organisations perform back office activities more efficiently."

The care tools are counted separately on the same page. The most widespread, the module that helps fill operating theatres, is counted in 40 trusts. The discharge tool is in 30. Cancer pathway tracking is in 6. None is close to 137.

80, then 137

Web archives kept the earlier versions of the page, and they show how we got here. At the end of March 2025, NHS England counted 72 trusts realising benefits and noted that these came from nationally commissioned products. End of September 2025: 77. End of February 2026: 80, and the note about products had gone.

The next update, published in June, gives 137 for the end of May. In three months, the category gained 57 trusts. That version is the first to carry the sentence about back-office tools.

Meanwhile the care-tool counters barely move. Operating theatres stays at 39, discharge at 25, and waiting list validation goes from 23 to 27. Those counters are published a quarter late, so we can't say the 57 newcomers are all back-office. But no care tool gained anywhere near that many hospitals.

It's like changing the rules mid-game: the score goes up, and nobody has scored a new goal.

Not a lie

The definition is written in black and white, and NHS England had already said as much: its list includes trusts that use the platform to analyse and submit data, not just to treat patients. According to the Guardian, some hospitals that say they don't use it may be using a small part of it to work out what treatments cost. Saving time on a cost spreadsheet is a real benefit, the way a well-fitted electricity meter is useful. It just isn't the one you picture when you read "137 hospitals benefit".

The problem lies elsewhere. For a care tool to be counted, the published method requires active use by clinical teams. For the headline figure, the one the agency sets against its critics, the methodology page describes no calculation. Asked by the Guardian, NHS England says nearly 140 trusts "have now reported benefits from the wide range of tools available on the platform".

These numbers have been scrutinised before. In July, the Office for Statistics Regulation (OSR) took note of a sentence NHS England had just added: the agency can't conclude that the platform caused the extra operations it reports. NHS England also committed to flagging any methodology change when it happens. The June widening came before that commitment. On the 29 hospitals, the OSR told the Guardian it is looking into the matter.

Even inside the NHS, there is unease. Marc Farr, who chairs the network of hospital data leads, says he is concerned and wants benefits reported accurately, transparently and verifiably.

What the number doesn't settle

Palantir defends its record with care outcomes: 117,000 extra operations, and discharge delays cut by 14.3% for long stays. Its UK and Europe chief, Louis Mosley, says ending the contract would be "putting politics above patients". Those results come from the care tools, which are deployed in around forty trusts at most.

In 2027, the government will have to say whether it keeps Palantir. The figure of 137 will tell it how many pipes have been laid. Not how many hospitals treat patients with them.