NHS Federated Data Platform

[Dame Siobhain McDonagh in the Chair]

I beg to move,

That this House has considered the NHS Federated Data Platform.

We are at a key time for the NHS as it changes from analogue to digital, and data is key to achieving better health results for all our constituents, as well as to the future of the NHS. AI analysis of scans can spot patterns of disease before the human eye, and modern communications can be much more effective than sending letters, which often arrive late. We are, however, at a junction where we can correct a series of mistakes made in the direction of travel in this process. I ask the Government to mind the gap between expectation and reality. We can and need to change.

In November 2023, a contract for services approaching £500 million was signed for the federated data platform. The Government’s contracts tracker describes a data platform owned and controlled by the NHS

“to unlock the power of NHS data to understand patterns, solve problems, plan services for local populations and ultimately transform the health and care of the people they serve.”

Sadly, the FDP developed by Palantir is far from that description.

The NHS is an inherently distributed organisation, with trusts in charge of their own IT. Although NHS England has been working on a unified data dictionary and standards, imposing a single central IT solution has yet to work. Indeed, a single central system can become a single point of failure. Such a critical element of national infrastructure must be under full control, fully owned and trusted.

Although I understand the appeal of a slick salesman who persuades that they can solve all the problems in the NHS, build that one system to bind them all, and use AI like magic to provide all the answers, sadly, it is not reality. Is Palantir’s FDP a product that the NHS can own and trust, or have we bought the emperor’s new clothes that, after huge investment, leave us with nothing? I will outline why this solution is wrong in three significant points: the contract is wrong, the solution is wrong and the supplier is wrong and simply not delivering on its promises.

I thank the hon. Member for securing this important debate at this critical point. Does he agree that the opaque procurement of the Palantir contract, one of Mandelson’s dodgy deals, is deeply concerning? Does he agree that the full details of Mandelson and the Prime Minister’s visit to the Palantir headquarters in 2025 must be made public?

I agree with the hon. Lady entirely. The secret meeting in 2019 between Boris Johnson and Dominic Cummings and Peter Thiel—the founder and chair of Palantir—that started this whole thing, for which there are no minutes, must be clarified as well.

I ask the Minister to consider using the contract renewal point to stop the chaotic expansion of the Palantir platform monopoly, to work to a staged exit with a retender for British companies to build a replacement for Palantir, and to deliver a better, long-term solution providing British sovereign capabilities in line with principles outlined by the Science and Research Minister and the Prime Minister.

The current contract delivers a subscription service that leaves no deliverables after the subscription—no software, no improvements and no intellectual property after spending more than £330 million. All the specially written software and intellectual property rights belong to the supplier, says the contract. All the rights to any know-how are explicitly retained by the supplier and not passed across on termination of the contract. The contract delivers no software—not one line—just a subscribed service; a permanent lock-in; a single point of failure.

Why are we building a leased service wrapped in glossy marketing promises, rather than a product that the NHS can own and trust? We are paying the supplier to hire Accenture, PwC, NHS experts and consultants to create a solution that we do not own—the supplier does. It uses external AI platforms from OpenAI and Anthropic and brings questionable value itself. Prior to it buying an opportunity to provide its system to help manage the data from the covid vaccine programme, the supplier had no expertise in health.

The three-year contract asks for 13 core capabilities to be delivered. According to the National Audit Office and the supplier, after nearly three years, it has partially delivered on three or four of those capabilities. Hon. Members may have received letters from the supplier, which has also taken to sponsoring newsletters that we see every day.

My hon. Friend is making an excellent case. I know that, in a moment, he will come on to the point that this contract is coming to an end. I am sure that it is being reviewed by the Government—the Minister will respond on that issue—but we are encouraging them to bring the contract to a close, for the reasons that my hon. Friend is properly explaining. He will perhaps also agree that we should go through a transition period to ensure that the conditions he has described are addressed, so that the Government can benefit from the software that has been developed.

I think my hon. Friend has been reading my speech in advance. I absolutely agree with him.

I see that the outgoing NHS England chief data and analytics officer, Ming Tang, has publicly joined Palantir’s fightback, saying that the system is delivering—but having introduced Palantir and lobbied to deploy it, she would say that, wouldn’t she? Given Palantir’s habit of lobbying civil servants and the revolving door from Government, I wait to see where she will end up.

I ask the Minister to review the contract, particularly in the light of the Government’s policies on investing in UK tech, value for money, technical lock-in, key performance indicators and strategic supplier status, which suppliers should have. I ask the Minister to reject extending the existing contract, which locks in the NHS forever and delivers nothing tangible.

My hon. Friend is making some important points. Just yesterday, I highlighted the Scottish Government’s decision to buy Chinese buses, which come with a serious security risk. It would be wrong of me not to do the same when the UK Government take the same risks. We have some of the finest minds in the world here in the UK, but too frequently, we lose them to foreign firms that are out of our control. Does my hon. Friend agree that we should be developing Britain’s skills base here, and that whether it be buses or AI, we should be putting British jobs and British security first?

I absolutely agree with my hon. Friend, as do the Prime Minister and the Science and Research Minister.

Turning to the solution, the solution is wrong. There have been many attempts to unify the NHS by using a single IT system; each one has failed. In reality, we must think of the NHS as thousands of independent organisations. NHS England has been guiding organisations towards a combined data dictionary for more than 10 years, combining definitions of what data means, how it is recorded and the way it is used. After three years, about half of the 200-odd NHS trusts across 42 integrated care boards are quoted as live on the FDP, and only a quarter of them report benefits from using it.

My hon. Friend’s expertise in this area is impressive. I recently spent a shift with the South Western ambulance service and saw how critical it was for that service to be able to access both GP and hospital data—we had a lady who had had a heart attack, and we did not know who she was. Does my hon. Friend agree that the priority should be for the different elements of the NHS to talk to each other, rather than be scraped by a third party such as Palantir?

I do agree. Palantir will not enable GPs, paramedics or anybody else to see hospital data. They will have to go through Palantir to see that data, and they will not be able to access patient records from the hospital to the GP or vice versa.

Like any data warehouse, Palantir requires connecting software that reaches into each of the NHS’s internal systems and gathers data. That data gathering is being done NHS trust by NHS trust, as there are differences inside each one. That is embedding the use of Palantir-owned code inside every NHS trust by creating custom connecting software to connect and translate data.

In Devon, the local ICB has celebrated as a major success the adoption of the same electronic patient record across Devon’s four main hospitals. It has just gone live in Torbay trust, which serves most of my Newton Abbot constituency. In an organisation as diverse as the NHS, with such distributed responsibility, we can either impose one massive system to rule them all or build interoperability. Interoperability would allow GPs to see hospital records and vice versa. Palantir is not doing that.

Interoperability is how massive systems, such as the internet or mobile phone networks, work. They do not rely on one single system or supplier. In that way, a modular system, a bit like Lego, can be constructed that, overall, is immune to changes elsewhere in the wider environment, providing only the specific data required to deliver improvements in services. That form of system builds long-term capability and delivers without requiring a locked-in, expensive subscription. It can also be built by a UK tech consortium in parallel with phasing out Palantir, which would build UK sovereign solutions, tech skills and competencies.

Meanwhile, NHS England’s October 2025 medium-term planning framework mandates all NHS providers of acute, community and mental health services to sign up to the FDP, and it demands that any existing local data analysis systems are removed. That results in further lock-in.

One of the issues that the hon. Gentleman has not discussed in his excellent speech is governance. The commissioner of the contract is the NHS, and it is also the main oversight body. He has put forward a black-and-white solution—end or maintain the contract—but is there not a case for a more robust governance structure? That could involve giving more powers to the National Data Guardian or setting up a bespoke oversight body for this contract.

If the right hon. Gentleman looks at the contract, he will see that it is subscription only: deliveries under supplier software—none; deliveries under third-party software—none. Any programming written under the contract is owned by Palantir. The contract has to be adapted for any next phase so that Palantir can be moved out.

Palantir is not only the wrong technical solution; NHS users report that it is awful to use. An open letter to NHS England said:

“we already have similar tools in use that presently exceed the capability and application of what the FDP is currently trying to develop or roll out”.

An NHS worker said:

“We’re being forced to use a convoluted system that makes even the simplest tasks feel like pulling teeth. It’s demoralising, and honestly, it’s a waste of everyone’s time and public money”.

An NHS data analyst said:

“Not only could similar functionality have been delivered at a fraction of the cost, but the existing tools are already better integrated, more intuitive, and more conducive to collaboration”.

In early 2025, Greater Manchester ICB reported that the FDP

“does not currently have any system-level products that offer the same or better functionality, compared to the custom-built system already in use for NHS GM”.

An NHS developer concerned about Palantir wrote to me to say:

“There are any number of reassuringly boring companies that could deliver this contract, many of them based in the UK, and then we could just get on with the exciting work of using technology to improve care for our patients”—

quite right too.

We must halt this path of chaos before the costs build any higher. I ask the Minister to use the change from NHS England to allow a change of direction towards a distributed, interoperable UK sovereign solution. Will the Minister cancel the expansion of the FDP to community and mental health service providers?

Order. Mr Wrigley has asked me if he can sum up at the end of the debate—his chances were in his hands. Because the debate is so popular, I will now impose a two-and-half-minute time limit on contributions. I call Dawn Butler.

It is wonderful to serve under your chairmanship, Dame Siobhain.

The reality is that the Government have inherited a mess. We must not take responsibility for that mess. I congratulate the hon. Member for Newton Abbot (Martin Wrigley) on securing the debate; he mentioned the £1 that Palantir paid to get its foot in the door. Some might say that was a smart business move; others might say that was a con or a trap. The first responsibility of a Government is to ensure that they protect their citizens. In order to do that, we must have AI and data sovereignty. We must ask ourselves: does Palantir allow us to have that AI and data sovereignty? I think the answer to that is no.

We have too much reliance on US systems. I was in hospital just the other week and the system is designed for the US, so that QR codes can be scanned at every action, because in America patients have to pay for everything, from blood tests to anything else. We do not have to deal with that, because luckily we have the NHS.

To say that we cannot do better in the UK just is not true. Unison and the BMA have severe concerns about Palantir, and they have said that they would like the Government to have a system that is based on ethics and values. That should be our starting point—a system of ethics and values. There are better alternatives and it is untrue to say that there are not. The BMA and Unison briefing talks about local platforms that already exceed FDP capability. For example, OpenSAFELY, developed at the University of Oxford, is an open source, privacy-preserving platform that has supported more than 200 NHS research projects.

The reality is that 30 people decided on the contract. Unison represents 1.3 million public service workers, and the BMA represents all doctors. They all have concerns. The Government need to cut loose the relationship with Palantir and have a transparent review that includes workers, trade unions and the public.

It is a pleasure to serve under your chairship, Dame Siobhain.

In a briefing for investors, Palantir chief executive officer Alex Karp said:

“we are super proud of the role we play, especially in places we can’t talk about…Palantir is here to disrupt…and when it’s necessary, to scare our enemies and on occasion kill them.”

If it looks evil, if it smells evil and if it behaves evil, then it is evil. I thank the hon. Member for Newton Abbot (Martin Wrigley) for securing this debate. Given my lack of speaking time, I am going to have to cut my speech short.

We must examine Palantir’s record elsewhere in the world in our assessment of its suitability as a supplier to the NHS. Palantir has long-standing contracts with United States Immigration and Customs Enforcement, and the tools to track and target migrants that it has provided to ICE have facilitated racial profiling, family separation and violations of due process. In January 2024, Palantir entered into a strategic partnership with the Israeli Ministry of Defence to support war-related missions, meaning that its advanced data systems are used to perpetuate a genocide against the Palestinian people. Reports from Novara Media and the Financial Times state that NHS staff have been warned that they could be fired for criticising the NHS contract with Palantir, that NHS organisations are under pressure to sign up voluntarily to the Palantir data system, and that technicians have been told to stop working on alternate systems.

The Swiss Government have previously rejected engagement with Palantir, raising national security concerns; in the US, hospital systems in New York have moved away from Palantir arrangements, citing data governance and control issues; and here at home, uptake of the FDP remains uneven, with many NHS professionals reluctant—in some cases refusing—to engage with it. Perhaps most concerning of all are the reports that patient data opt-outs do not apply to the federated data platform due to a legal direction issued under section 254 of the Health and Social Care Act 2012, yet that direction has not been published or named or been subject to scrutiny. This debate is not about technology; it is about who we trust to sit at the heart of our NHS, and on that question, the Government must think again.