NHS Palantir FDP Refusal: When Corporate Ethics Become a B2B Adoption Barrier

A £330 million contract. Seven years of guaranteed revenue. A consortium led by one of the world’s most powerful data companies. And yet, NHS staff are logging into Palantir’s Federated Data Platform and feeling physically ill.

“It makes me feel sick every time I log into the thing,” one NHS official told The Register in April 2026. Another called the platform “ethically bankrupt.” Across England’s hospitals, a quiet revolt is unfolding — not against bad software, but against the company that built it.

The NHS Palantir FDP Refusal Is Not a Bug — It’s a Feature of Ethical Failure

This is the story of “software refusal” — a phenomenon where end users deliberately boycott, slow down, or circumvent enterprise tools not because the tools don’t work, but because the vendor’s reputation makes using them feel morally untenable.

Why a £330M Platform Can’t Buy Trust

The NHS Palantir FDP refusal represents something unprecedented in enterprise technology: a product that passes every procurement metric — on budget, on schedule, technically functional — yet faces systematic rejection from the people who are supposed to use it.

The implications extend far beyond the UK’s National Health Service. For any enterprise buyer evaluating vendors with controversial histories — military contracts, surveillance work, human rights complaints — the Palantir case is a warning: ethics is not a PR problem. It is a structural adoption barrier.

Background: Palantir’s Road From CIA to NHS

Palantir Technologies was founded in 2003 with seed funding from In-Q-Tel, the CIA’s venture capital arm. The company’s name comes from the seeing stones in Tolkien’s Lord of the Rings — a fitting metaphor for a firm built to see everything.

Co-founded by Peter Thiel and CEO Alex Karp, Palantir spent its first decade as an intelligence and military contractor. Its Gotham platform helped the CIA and US Department of Defense track targets. Its Foundry platform expanded into commercial markets. The company went public in September 2020 at a $22 billion valuation.

But Palantir’s most controversial work preceded its commercial pivot. In 2014, the company signed a $41 million contract with US Immigration and Customs Enforcement (ICE) to build the Investigative Case Management system — a platform that processed medical records, phone logs, and location data to track immigrants.

In April 2025, ICE signed a new $30 million contract with Palantir for near real-time migrant tracking capabilities. By June 2025, Palantir’s CTO Shyam Sankar was sworn in as a US Army Reserve lieutenant colonel — a move that blurred the line between tech company and military institution.

Then came the quote that crystallized everything. During Palantir’s Q4 2024 earnings call in February 2025, CEO Alex Karp stated: “Palantir is here to… when it’s necessary, to scare enemies and on occasion kill them.”

The NHS Federated Data Platform Contract

In November 2023, NHS England awarded the Federated Data Platform (FDP) contract to a Palantir-led consortium that included Accenture, PwC, NECS, and Carnall Farrar. The deal: £330 million over seven years, covering up to 240 NHS organisations.

The FDP’s purpose is straightforward — integrate operational data across NHS trusts to manage waiting lists, patient records, bed capacity, and resource allocation. On paper, it solves a genuine problem: the NHS has historically operated on fragmented, incompatible data systems.

NHS England targeted 85% trust adoption by March 2026. The trajectory tells the story: Corporate Watch’s FOI requests found only 34 trusts actively using FDP in early 2025. By October 2025, that number rose to 77; by January 2026, 110 trusts were live. But even 110 out of 205 trusts falls far short of the 85% target — and “live” does not mean “fully adopted.”

The NHS Palantir FDP Refusal Anatomy: How Staff Revolt Against Approved Software

The gap between official metrics and ground-level adoption tells the real story. Even as the numbers improved on paper, hospital corridors told a different story. Staff at trusts classified as “live” described workarounds to avoid using FDP — logging in to satisfy compliance checks, then reverting to legacy systems for actual work.

This pattern — steady on-paper progress masking ground-level resistance — is not a rollout failure in the traditional sense. The software works. The infrastructure exists. Staff simply refuse to engage.

The BMA’s Formal Rejection

On January 27, 2026, the British Medical Association published its position in the British Medical Journal: doctors should “immediately take steps to explore refusing any non-direct care usage” of FDP, with a view to “moving away from the platform entirely.”

The BMA’s rationale centered on Palantir’s ICE contracts — specifically the use of medical records in immigration enforcement. For a medical union whose members take an oath to protect patient welfare, the connection between a healthcare data platform and immigrant surveillance was intolerable.

This was not a fringe position. The BMA represents over 190,000 doctors across the UK. Its recommendation carried institutional weight.

Amnesty International’s Campaign

In March 2026, Amnesty International UK launched the “No Palantir in our NHS” campaign, a joint briefing with healthcare advocacy group Medact. The campaign urged all NHS trusts and Integrated Care Boards to refuse FDP implementation.

Amnesty’s argument went further than the BMA’s. It cited Palantir’s role in Israeli military operations in Gaza, its predictive policing work (dating back to New Orleans in 2012), and its broader surveillance infrastructure. Medact warned that FDP health data could be linked to other Palantir platforms to enable US-style immigration raids in the UK.

A critical detail: FDP implementation is not mandatory. Individual trusts and ICBs can decline. This gave the campaign a concrete, actionable target — persuade local decision-makers, one hospital at a time.

The “Software Refusal” Phenomenon

The Register’s April 3, 2026 report documented what frontline staff were actually doing. NHS workers were:

• Deliberately slowing down workflows involving FDP
• Using alternative tools to bypass the platform
• Outright boycotting the system
• Expressing moral distress (“ethically bankrupt,” “makes me feel sick”)

This is “software refusal” — a form of workplace resistance that inverts the typical enterprise IT dynamic. Normally, adoption resistance stems from usability issues, training gaps, or change fatigue. Here, the resistance is moral. The tool works fine. The vendor does not pass the ethical bar.

Parliamentary and Political Pressure

The pressure escalated to Parliament. On February 10, 2026, MPs debated Ministry of Defence Palantir contracts in the Commons. On March 10, 2026, a broader Technology Sovereignty debate addressed Palantir’s role in UK public infrastructure.

Green Party MP Zack Polanski stated Palantir should “pack its bags and get the hell out of the NHS.” Chi Onwurah, chair of the Science and Technology Select Committee, called concerns about the contract “legitimate and substantive.”

The government has sought legal advice on triggering the contract’s break clause, which comes up in February 2027.

International Contagion: NYC Hospitals Drop Palantir

The NHS was not alone. In March 2026, New York hospitals ended their Palantir contracts amid the same ethical scrutiny. The contagion effect confirmed that this was not a UK-specific political issue but an emerging pattern in healthcare IT procurement globally.

The NHS Palantir FDP Refusal Paradox: When Good Software Meets Bad Reputation

Palantir UK Vice President Louis Mosley dismissed the opposition as “ideologically motivated attacks.” From a procurement standpoint, he has a point — the FDP met its budget and timeline benchmarks. The software functions as specified.

But this defense reveals a fundamental misunderstanding of enterprise adoption. Procurement and adoption are different processes. Procurement involves committees, RFPs, and evaluation criteria. Adoption involves humans deciding to open an application every day.

The NHS Palantir FDP refusal introduces a new category: ethical adoption barriers. These bypass traditional procurement gatekeepers entirely. A vendor can win every competitive evaluation and still lose on the ground floor.

Consider the economics. NHS England committed £330 million for a platform that, at current adoption rates, serves 34 of 205 trusts. That is roughly £9.7 million per actively engaged trust — an extraordinary cost-per-adoption figure for a platform that was supposed to achieve near-universal rollout.

The principal-agent problem is stark: the entity that signed the contract (NHS England central procurement) is not the entity that uses the product (individual trust staff). When the users have moral objections that the buyers did not anticipate, the entire investment is at risk.

Korea Lens: Lessons for Public Sector Tech Procurement

Korean enterprises and public institutions face a version of this problem that is only beginning to surface. As government agencies and hospitals adopt foreign enterprise platforms for cloud, AI, and data management, the ethical track record of vendors is rarely part of the procurement evaluation.

The Palantir case offers three specific lessons for Korea’s public sector:

Ethical Due Diligence Must Become a Procurement Criterion

The Korean government’s public procurement system evaluates technical specifications, cost, and vendor financial stability. It does not systematically evaluate a vendor’s human rights record, military contracts, or surveillance history. The NHS Palantir FDP refusal experience shows this gap can invalidate an entire deployment.

“Software Refusal” Can Happen Anywhere Professionals Have Moral Authority

Korean doctors, teachers, and civil servants have professional identities that extend beyond their job descriptions. If a major Korean hospital adopted a data platform from a vendor linked to surveillance or human rights abuses, the same dynamic could emerge — especially given Korea’s strong labor unions and professional associations.

Data Sovereignty Is About Who Handles It, Not Just Where It’s Stored

Korea’s Personal Information Protection Act (PIPA) regulates data handling, but it does not address the ethical profile of data processors. The FDP case suggests that public trust in healthcare data systems depends not just on compliance, but on the perceived integrity of the vendor.

Korean B2B buyers evaluating foreign enterprise vendors — particularly in healthcare, defense, and public administration — should treat the Palantir precedent as a case study in procurement risk. The cheapest, most technically capable vendor may carry reputational liabilities that make adoption impossible.

Conclusion: The Palantir Precedent

The £330 million NHS Federated Data Platform contract may ultimately die not from technical failure, budget overruns, or competitive displacement. It may die because the people who were supposed to use it decided that the company behind it was morally unacceptable.

This is the Palantir Precedent: in enterprise B2B, your other clients are your reputation. When those clients include ICE, military intelligence agencies, and operations that trigger Amnesty International campaigns, no amount of technical excellence can overcome the trust deficit.

For enterprise vendors, the lesson is existential. Ethics is not a communications department problem. It is a revenue problem. Every military contract, every surveillance deal, every CEO statement about “killing enemies” becomes part of your sales pitch — whether you want it to or not.

For enterprise buyers — in the UK, Korea, and globally — the lesson is practical. Procurement must include ethical due diligence, or organizations will face the same adoption gap that turned a £330 million investment into a 17% active usage rate.

The bigger question remains: who should manage a nation’s healthcare data? The answer, increasingly, is not a company whose CEO talks about killing people.

References

1. “NHS staff resist using Palantir software,” The Register, April 3, 2026. Link
2. “The Palantir Problem: NHS Staff Are Quietly Refusing,” WebProNews, 2026. Link
3. “BMA calls for NHS doctors to reject using the FDP,” Digital Health, February 2026. Link
4. “Doctors told to pull back from Palantir’s NHS data platform,” The Register, February 11, 2026. Link
5. “No Palantir in our NHS,” Amnesty International UK, March 2026. Link
6. “FOI requests reveal Palantir’s NHS FDP rollout failures,” Corporate Watch, 2026. Link
7. “Government right to review NHS Palantir deal, says MP,” Computing.co.uk, 2026. Link
8. “MoD Palantir Contracts Debate,” Hansard, February 10, 2026. Link
9. “Palantir’s Billionaire CEO Just Can’t Stop Talking About Killing People,” Gizmodo, 2025. Link
10. “What is Palantir?,” American Friends Service Committee (AFSC). Link
11. “New York hospitals drop Palantir,” ResultSense, March 2026. Link
12. “Palantir FDP Health Briefing,” Medact, 2026. Link
13. “The Palantir-NHS Partnership,” LSE Media Blog, July 2024. Link

Related reads on The ByteDive:

• Iran War and AI Supply Chain Paradox — How geopolitical conflict reshapes tech supply chains
• 16 Million Stolen Queries: Nation-State AI Weaponization — Data sovereignty in the age of AI
• AI Capital’s 3-Layer Shift — Why ethical risk is becoming a new layer in AI investment evaluation