The NHS has invested over a billion in software it has no power to deploy

How student politics in council chambers is blocking national health infrastructure

Britain cancelled the northern half of HS2. It takes fifteen years to approve a runway extension. Housing targets are missed every year because local opposition defeats national need. Organised opposition, however small, consistently defeats silent majority support for infrastructure the country needs.

That pattern has now reached the NHS.

NHS England has a policy position: FDP first. Every trust should be using the Federated Data Platform, the national data infrastructure connecting clinical and operational data across 220 trusts. This includes a planning document that explicitly states Trusts should adopt FDP as their data warehouse. And yet a council with no authority over NHS procurement seems to have found a way to block it. The government seems happy with this.

Sheffield council has passed a symbolic motion, thereby successfully pressuring trusts through scrutiny committees, and the national programme has no mechanism to stop it. At Doncaster and Bassetlaw Teaching Hospitals, an FDP project was killed internally and the trust will now spend £600,000 of public money on an inferior alternative. I am told the decision arose due to local politics, and the summary reason was "Palantir = Bad".

If that pattern repeats, the cost is staggering. Half a million pounds per trust in avoidable local data warehouse costs amounts to over £110 million across the NHS. That does not include the associated managed services that eat away at the Trust's revenue budgets. It also does not account for the billions in lost potential from a fragmented landscape where every trust builds alone, no clinical tool is portable, no data is standardised, and the AI infrastructure that every trust board is being told to prepare for has no foundation to build on.

In June, Sheffield City Council became the first local authority to formally oppose FDP. The motion passed 62-1. That sounds decisive until you look into the detail and realise how superficial the discussion around the motion was. Across three meetings, five months, and more than twenty speakers, not one person discussed what the platform does. The debate was entirely about Palantir, and was littered with inaccuracy and a disregard for patients' or taxpayers' interests.

A campaign that wrote its own evidence base

In March, the Health Scrutiny Sub-Committee invited NHS trusts and the South Yorkshire ICB to discuss FDP. None attended. Campaign groups did. Medact Sheffield presented unopposed. Councillors heard about ICE raids, genocide, and surveillance capitalism. They did not hear about discharge coordination, cancer pathway tracking, referral triage, or any of the products clinicians use daily. A scrutiny committee that hears only from one side is not conducting scrutiny. It is receiving a brief.

The committee unanimously expressed concern and made FDP a standing agenda item. One councillor stated the council would not "touch Palantir because the UN has labelled it as a company to avoid." The UN has done no such thing, in fact the UN World Food Programme actually uses Palantir's software. The claim was not challenged.

Other claims were wrong. "Palantir has used US medical data to target minorities for deportation" misattributes the decision. US government agencies decided to share Medicaid data with immigration enforcement. Palantir built the tool they used. The decision on what data to share was the government's, not the supplier's. Nobody blames the manufacturer of a filing cabinet for what the government puts in it. "Greater Manchester rejected FDP for being inadequate and poor value" is contradicted by Manchester's own board report, which acknowledged FDP may come to exceed some of their platform's capabilities.

The committee then did something remarkable. It voted unanimously to formally record every member's concern about Palantir's suitability, based entirely on the testimony it had just heard from campaign groups with no NHS organisation in the room to respond.

It then voted to send the campaign testimony, the public statements, and the meeting minutes directly to Sheffield's NHS trusts and the South Yorkshire ICB as formal committee correspondence. The committee did not just hear the campaign. It became the campaign's distribution channel, forwarding one-sided material to NHS organisations as if it were scrutiny evidence.

The council became a circus

By June, the full council debated. Ten councillors spoke. One compared Palantir to a cursed artefact from Lord of the Rings. Another invoked Aneurin Bevan. The Labour amendment, formally moved by Councillor Ruth Milsom, described Palantir's UK head as "the grandson of Oswald Mosley" and praised Greater Manchester for rejecting FDP as inadequate and poor value. That ancestry reference was not an off-the-cuff remark. It was written into the formal text of an amendment that passed 61-1. A third councillor warned that Palantir could enable a future government to weaponise health records for mass deportations, as though that would not be possible if another supplier were appointed.

Perhaps the most extraordinary claim came from Liberal Democrat Councillor Barbara Masters, who told the chamber that Palantir's software allows the company to compile and analyse databases across government departments in order to "influence the decisions that the government makes." That is a claim that a data processing supplier is using its clients' data to manipulate the governments it serves. No evidence was offered. None was asked for. If a company that builds database software can be accused in a council chamber of puppeteering the British government, with no challenge from the floor and no consequence for the person making the claim, the quality of scrutiny that preceded this vote should concern everyone who relies on the NHS technology decisions that flow from it.

But amidst the nonsense, one absurdity is true: adoption of FDP has not been formally mandated, and so local organisations believe they can procure whatever systems they want regardless of national direction.

NHS England has set a clear direction of travel. Every trust and ICB is required to submit a plan within two years detailing how they intend to adopt and use the platform. The use of FDP products has been mandated in NHS England's operational planning guidance. Sheffield Children's own written submission to the scrutiny committee confirmed this.

But the government has stopped short of making adoption compulsory, leaving a gap wide enough for any council motion, any campaign group, and any trust board looking for an excuse to drive through. The result is a national programme costing over a billion pounds that can be undermined by a symbolic vote in any council chamber in the country. That is a structural failure the government needs to urgently rectify.

Scrutiny conducted by the prosecution

Two weeks later, the scrutiny committee reconvened, now chaired by Councillor Ruth Milsom, the same Labour councillor who had moved the amendment opposing FDP. The person leading scrutiny of trusts' adoption decisions had already voted to block the platform.

Again, no trust representative attended. Written submissions were read by a council officer. Sheffield Children's had confirmed it would proceed with FDP for daycare scheduling. The only substantive technical question of the entire scrutiny process came from Reform UK Councillor Jack Byron, who asked whether trusts running different platforms would be able to share data and what the impact on patients would be. It was the right question. The chair's response was not about data: "patient trust is already eroded so far in Palantir that I don't think any of us can see how it's going to work across the system." She then added, "that's from a non-data expert point of view"' The chair of the scrutiny committee acknowledged she was not qualified to answer the question and then answered it anyway with a predetermined conclusion. That is a verdict, not scrutiny, and her opinion extends to the "system" by which she means the whole country.

The committee agreed to summon Sheffield Children's Hospital to justify its decision to press ahead with FDP adoption. Read that again. The hospital was called in to explain why it was following a nationally commissioned project which has been set out as a must do in the NHS planning guidance. The press reported "Sheffield Children's Hospital bosses questioned over Palantir."

The ICB's own committee papers now formally record the damage: "There is a risk that clinicians, staff and the public's concern about the use of the Federated Data Platform leads to reduced trust in how the ICB uses health data, this may result in reputational harm and resistance to sharing data that enable us to inform population health management." The political campaign is now documented by the ICB as a risk to population health management in South Yorkshire. This was not public concern, it was a campaigner manipulation dressed up.

The pattern spreads

Rotherham Council tabled a similar motion three weeks later. Rotherham Friends of Palestine attended. The motion names both Rotherham NHS Foundation Trust and Rotherham, Doncaster and South Humber NHS Foundation Trust. The Sheffield model is replicating across South Yorkshire, council by council, each with the same structure: campaign groups present, trusts absent, votes passed, clinical decisions pressured.

The pressure is also flowing through NHS governance structures. Sheffield Teaching Hospitals told the scrutiny committee it is not adopting FDP, citing the implementation of a new EPR. Doncaster and Bassetlaw Teaching Hospitals, which approved a business case in 2025 to share Sheffield Teaching Hospitals' Oracle Health EPR instance and sits in the same South Yorkshire Acute Federation with board-to-board governance, subsequently killed its own FDP project and is spending £600,000 of public money on an alternative. The decision was attributed to local politics. What began as a symbolic council motion is now transmitting through formal NHS structures into clinical technology decisions at neighbouring trusts.

Every council in the country commissions services that depend on national infrastructure it did not choose. The electricity grid runs on technology procured nationally. The water supply is managed by regulated companies the council did not select. Nobody tables a motion demanding the local hospital stop using the electricity because they object to the generator manufacturer.

The FDP is national health data infrastructure. It was procured through a lawful process, awarded by NHS England, and deployed across all trusts. The break clause decision is rightly a matter for ministers, informed by evidence, cost-benefit analysis, and the Imperial College evaluation. It is not a matter for council chambers where the debate contains factual errors, no technical scrutiny, and comparisons to fantasy novels.

The cost is paid by patients

What happened in Sheffield is populism applied to clinical infrastructure. A campaign group briefed a committee with one-sided testimony. The committee passed a motion without discussing what the technology does. The scrutiny process was chaired by someone who had already voted against the platform. The facts presented were irrelevant and inaccurate. Tragically, the resulting pressure is costing trusts public money, silencing practitioners, and degrading the ICB's ability to manage the health of its population.

The patients who benefit from the tools that FDP provides will continue to be managed on paper, spreadsheets, and phone calls while councillors who cannot describe what the platform does decide whether their local hospital is allowed to use it. If you are a patient, your care will be weaker than it would have been if Doncaster had gone ahead with FDP as its warehouse. If you are a taxpayer you are now diverting £600k plus revenue costs to an alternative no one needs. And if you are a voter then your government has declined to give the NHS the power to drive through improvements it desperately needs. That needs to change.