NHS Patients Can't Opt Out of Palantir. Their Hospitals Can.

The Opt-Out Asymmetry: What the Minister Actually Said

The asymmetry is the cleanest possible statement of who has choice and who does not in the NHS Federated Data Platform, and it was confirmed on the floor of the House of Commons on June 13, 2026 by the health minister responsible for the FDP.[1] Preet Kaur Gill, who was appointed last month to cover health innovation and safety, told Labour MP Neil Duncan-Jordan that patients can only opt out of "secondary uses" of NHS data, such as planning and research. On the main opt-out mechanism, the National Data Opt-Out, the minister said: "The National Data Opt-Out does not currently apply to products used in the NHS FDP. In most cases, this is because data is being used for the purpose of direct care."[1]

That phrasing is the policy line that the NHS has been advancing since the contract was awarded to Palantir in November 2023, and it is the line the privacy and patient-advocacy community has been contesting since the same month.[4] The justification is that the FDP is a "direct care" platform, used by clinicians treating individual patients, and the National Data Opt-Out is, by policy, reserved for "secondary uses" of NHS data such as research and planning. The privacy-community counter-argument is that anonymization of NHS patient data is technically almost impossible, because the density of fields in a health record is high enough that even supposedly de-identified records can be re-linked to a named individual.[4] Conservative MP David Davis used exactly that framing when the contract was first awarded in 2023, and the same framing has been carried by the Foxglove legal campaign, the Doctors' Association UK, and the Medact briefing coalition throughout 2025 and 2026.[4][5][6]

The new piece on June 13, 2026 is the answer to MP Rachael Maskell's separate question: yes, NHS trusts running hospitals, mental health, and other services can opt out. The minister's exact words, again on the floor of the House: "Where NHS organizations would like to use alternative solutions, they retain the ability to procure locally, provided solutions meet applicable standards and support the delivery of national priorities."[1] That phrasing is a policy escape valve. It allows the Department of Health and Social Care to claim patients have not been denied a choice (because patients cannot opt out of "direct care" by design) while acknowledging that the institutional buyers can, in fact, walk away. The asymmetry is the design.

The Numbers: 168 of 214 Trusts, 123 Live, One ICB Holding Out

The actual deployment of the FDP is now close to saturation in England, and the saturation makes the opt-out asymmetry the dominant privacy question. According to NHS England statistics cited in The Register's June 13, 2026 report, 168 of 214 NHS trusts have signed up to use the Federated Data Platform, 123 trusts are live on the system, and 80 trusts are reporting measurable benefits from the rollout.[1] All but one of England's 42 integrated care boards (ICBs) have also joined. The lone holdout is Greater Manchester, the UK's second-largest health region, covering 2.8 million people and one of the most data-mature ICBs in the country.[2]

Greater Manchester's decision to refuse to sign the FDP is the most consequential single trust-level opt-out in the system, because it sits at the ICB layer rather than the trust layer. ICBs are the statutory NHS bodies responsible for planning and funding most hospital, mental health, and community services in their area; an ICB-level opt-out cascades to every trust and provider in the region. Greater Manchester's own reasoning, made public through a Freedom of Information response, was that "public concerns have heightened rather than diminished" since the ICB first deferred a decision in 2025, and that "there does not appear to be any compelling evidence that the value proposition for NHS GM from FDP has materially changed in favour of adoption."[2] The ICB also said its in-house data-analytics capability exceeded what the FDP currently offers, a position that directly contests Palantir's value claim on its own terms.[2]

The 168 of 214 trust number is the denominator question the privacy and patient-advocacy community has been asking since the contract was awarded. Of 214 trusts, 46 have not signed up. The Register does not name the other 45, but Greater Manchester is the only ICB to have publicly refused, and the most plausible reading is that the other 45 are a mix of trusts that have not yet activated the FDP and trusts that have, like Greater Manchester, declined on opt-out or capability grounds.[1] The asymmetry means that every patient in a trust that has signed up is currently having their data processed by Palantir, with no patient-level opt-out, and the only way to exit is for the entire trust to exit.[1]

The SIT Committee Report: One Week Earlier, "End the Contract"

The opt-out asymmetry is the latest in a sequence of escalating signals against the Palantir contract. On June 3, 2026, the House of Commons Science, Innovation and Technology (SIT) Committee published a report explicitly calling on the government to cut its ties with Palantir.[3] The committee described Palantir as "the most concerning example of the public sector's growing reliance on a small number of major technology providers," and recommended that the government use the February 2027 break clause in the FDP contract to "develop an in-house replacement or seek an alternative developed by UK-owned and UK-based providers that are more compatible with UK values, and do not pursue either technical or contractual dependencies."[3]

The committee's report was not a one-off. The chair, Dame Chi Onwurah MP, said in the launch statement: "A critical part of this transformation should include reducing the UK's dependence on a small number of big US tech companies like Palantir. Vendor lock-in isn't inevitable, and the current position leaves us seriously exposed."[3] The committee flagged Palantir's origins in US security, immigration, and defense contracting, and called out the political positions of Palantir co-founder Peter Thiel and CEO Alex Karp, including Thiel's stated view that "democracy and freedom are not compatible."[3] The committee was careful to say the recommendation was not ideological, but the language about "UK-owned and UK-based providers that are more compatible with UK values" is the strongest public-sector statement yet on Palantir's structural role in the UK state.

The report lands in a wider context. The Register's June 3, 2026 report on the committee publication also notes that MPs have tabled 40 written questions about Palantir in the last month, that the Department of Health and Social Care is the named buyer in the FDP contract, and that the UK government has spent the last year signaling a digital-state agenda that is, on the committee's reading, structurally incompatible with vendor lock-in to a single US supplier.[3] The committee's "reduce dependence on a small number of big US tech companies" framing is the bridge between the health-specific FDP question and a wider UK industrial-policy question that will run through 2026 and into 2027.[3]

The 2026 Privacy Shock: NHS England's "Admin Role" Change

The opt-out asymmetry is the most visible privacy question, but the most consequential 2026 development is a quieter one. Last month, NHS England confirmed, in response to a separate set of written questions from MPs, that it had changed the FDP policy so that some Palantir staff can access identifiable patient data through a new "admin" role.[1] A briefing document seen by the Financial Times and confirmed by The Register warned that granting that access could create a "risk of loss of public confidence" in NHS England's assurances about safeguarding patient data.[1] The Register's own coverage of the briefing is dated June 13, 2026, and the briefing was prepared by NHS England for internal use before being seen by the FT and The Register.[1]

The "admin role" change is the practical piece that makes the opt-out asymmetry matter in 2026 specifically. If patients cannot opt out, and Palantir staff can now access identifiable data through an admin role, then the operational gap between the public-facing "direct care" framing and the actual data-access structure inside the FDP has widened.[1] NHS England's stated safeguard is that the admin role is oversight and not direct clinical access, but the briefing document The Register obtained flags the safeguard as a public-confidence risk, not a technical one.[1] The point is not whether the safeguard works in practice. The point is that the policy was changed without the parallel public-facing communication that would let a patient know that the "direct care" framing is no longer a full description of the data-access structure.[1]

The "admin role" change is also the proximate trigger for the volume of written questions in the last month. The Register's June 13 report cites 40 written questions about Palantir tabled in the last month, with the breakdown tilted toward the FDP and the admin-role change in particular.[1] The political economy of the FDP is now running in two registers at once: a public-facing one in which Palantir is a "direct care" platform, and an internal one in which Palantir is a contractor whose staff have identifiable-data admin access. The opt-out asymmetry is the public-facing register's policy answer, and the briefing leak is the internal register's policy problem. The two registers have not yet been reconciled in public.

Why Greater Manchester Is the Holdout

Greater Manchester ICB is the structural test case for the trust-level opt-out, and its reasoning is the only public, on-the-record statement of why a competent NHS body would refuse the FDP.[2] The ICB manages health services for 2.8 million people, deferred a decision in 2025, and then confirmed in late 2025 that it would not reverse course because, in the ICB's own words, public concern has "heightened rather than diminished" and the FDP's value proposition has not materially improved.[2] The ICB also said its own data-analytics capability exceeded what the FDP currently offers, a position that is technically defensible for an ICB that has been investing in local data infrastructure for years.[2]

The Greater Manchester case matters because it shifts the burden of proof. If a single ICB can demonstrate that a trust can refuse the FDP and still deliver care for 2.8 million people, the policy escape valve Preet Kaur Gill confirmed on June 13 becomes a working precedent, not a paper one.[1][2] The next ICB to consider refusal has a Greater Manchester-shaped template to cite, and the next 45 trusts that have not yet signed up have a published statement from a same-size ICB that explains how they did it. The opt-out asymmetry is the rule; Greater Manchester is the test case for the exception.[2]

The other reading of the Greater Manchester case is that an ICB is structurally not the right unit of refusal. The ICB is a planning and funding body, not a direct care provider; the trusts that actually run hospitals, mental health services, and community clinics are the units that handle identifiable patient data. An ICB can decline to sign the FDP at the planning layer, but the trusts within its footprint could still individually decide to sign. The Register's report does not specify whether Greater Manchester's refusal is binding on its constituent trusts, and the FOI response The Register cites does not name the trusts.[2] The honest reading is that the ICB-layer refusal is real and consequential, but the trust-layer opt-out has not yet been exercised at scale. The asymmetry is still the rule; Greater Manchester is the proof that the institutional escape valve exists, not the proof that it is widely used.

The International Pattern: Palantir's 60-Day Losing Streak

The NHS FDP is not an isolated deployment. It is one of three or four major Palantir contracts that are under active legal, political, or commercial pressure in 2026, and the pattern across the last 60 days is the same: institutional buyers and oversight bodies are pushing back on Palantir's structural role in the public sector and the courts are agreeing.[7][8][9] The Financial Times reported on June 12, 2026 that Palantir lost a legal challenge against a Swiss investigative magazine that had published on the company's data-handling practices; the loss is Palantir's first significant international legal defeat in a friendly-jurisdiction press fight, and it lands two weeks after the New York City public hospital system announced it would not renew its $4 million Palantir contract.[7][8] The NYC hospital non-renewal followed public pressure, organizing by community groups, and a city-level review of the data pipeline that would have let ICE-adjacent contractors access patient records.[8]

The Swiss court loss, the NYC hospital non-renewal, the SIT Committee report, and the trust-level opt-out asymmetry are four different pressure vectors on the same company, in four different jurisdictions, in a 30-day window. The pattern is the story. Palantir's public-sector contracts have, for most of the last decade, been insulated by the framing that they are "operational efficiency" tools, that the data they process is sufficiently anonymized, and that the alternative is a slower or less capable public sector. The 2026 cycle is the first time that framing has been broken in four separate jurisdictions in the same month, and the institutions doing the breaking are the public bodies that were supposed to be the buyers and the courts that were supposed to be the backstop.[7][8][3][1]

The Palantir side of the response has been consistent: the company has not contested the substance of any of the four pressure points, and has instead emphasized the operational, technical, and counter-factual claims (the FDP is "on track" per the National Infrastructure and Service Transformation Authority, the contract "supports the delivery of national priorities," and a hypothetical alternative would "take planning, time, and resources to run a compliant procurement and then move services and data across safely").[1] The technical claims are not false, but they are not the question. The question is whether the opt-out asymmetry, the vendor lock-in, the human-rights record of Palantir's other public-sector customers, and the briefing-leak public-confidence risk are compatible with the "direct care" framing. The 2026 cycle is the first time the question has been asked institutionally, in four jurisdictions, at the same time.

What It Means for NHS Patients, Trusts, and Privacy Advocates

Three groups are affected by the opt-out asymmetry, and the consequences diverge sharply.

If you are an NHS patient in England. The opt-out asymmetry means that, in practice, you cannot prevent your NHS data from being processed by Palantir's FDP if the trust or ICB treating you has signed up. The 168 of 214 trusts that have signed up cover the vast majority of England's population.[1] The National Data Opt-Out, the long-standing patient opt-out for NHS data, does not apply to the FDP because NHS England classifies the FDP's processing as "direct care."[1] The privacy advocate position is that this classification is policy choice, not technical necessity, and that the only way to opt out as a patient is to opt out of NHS treatment at the trust that has signed up. The Department of Health and Social Care's position is that direct care processing is a separate category from secondary-use processing, that the opt-out was designed for the secondary-use category, and that the FDP does not change the patient's relationship with the clinician treating them.[1]

If you are an NHS trust board member, an ICB chair, or a chief information officer. The trust-level opt-out is the policy escape valve the minister confirmed on June 13.[1] Greater Manchester's ICB-level refusal is the published template for how to exercise it: defer a decision, gather public-feedback evidence, document the value-prop gap, and refuse on the published record.[2] The trade-off is real; the FDP is operational, and a refusal means building or buying an alternative. The minister's framing was that any alternative "would take planning, time, and resources to run a compliant procurement and then move services and data across safely."[1] The Greater Manchester answer is that the ICB already has the in-house capability and that the FDP does not currently offer enough incremental value to justify the privacy and political cost.[2] The trust-level opt-out is a working precedent, not a paper one, but it requires the trust or ICB to be ready to operate the alternative.

If you are a UK privacy advocate, a journalist covering the FDP, or a member of a campaign group such as Foxglove, Medact, or the Doctors' Association UK. The 2026 cycle is the first time the opt-out asymmetry, the SIT Committee report, the admin-role change, the briefing leak, and the trust-level opt-out have all been on the public record in the same 30-day window. The campaign-infrastructure argument now has more published primary sources than at any point in the FDP's three-year history.[1][3][4][5][6] The next 12 months, running up to the February 2027 contract expiry, are the procedural window in which the FDP's structural posture (Palantir as the FDP supplier) is set, regardless of which party is in office. The campaign that won the NYC hospital non-renewal was, in the end, a 14-month organizing and pressure campaign built on published FOI responses, public testimony, and elected-official pressure.[8] The UK campaign now has the equivalent building blocks, including the SIT Committee report, the 40 written questions, the briefing-leak documentation, and the Greater Manchester template.[3][1][2]

The Bottom Line

On June 13, 2026, health minister Preet Kaur Gill confirmed to the House of Commons that individual NHS patients in England cannot opt out of having their data processed by Palantir's Federated Data Platform, but individual NHS trusts can.[1] The National Data Opt-Out does not currently apply to FDP products because the data is processed for "direct care," and trusts "retain the ability to procure locally" if they want to use a different system.[1] The asymmetry lands one week after the House of Commons Science, Innovation and Technology Committee called on the government to use the February 2027 break clause and end the Palantir contract entirely, recommending an in-house replacement or a UK-owned alternative.[3] The asymmetry also lands the same week that NHS England confirmed, in response to separate written questions, that some Palantir staff can access identifiable patient data through a new "admin" role, a change a leaked NHS England briefing flags as a "risk of loss of public confidence" risk.[1]

Of England's 42 integrated care boards, only Greater Manchester has refused to sign up, citing heightened public concern, a lack of compelling value evidence, and an in-house data capability that exceeds what the FDP currently offers.[2] The trust-level opt-out is a real policy lever, and Greater Manchester is the published template for how to use it. The asymmetry is the rule, the trust-level opt-out is the exception, and the next 12 months (running to the February 2027 contract expiry) are the procedural window in which the structural posture of the FDP will be set.[1][3]

Watch for three things over the next 30 days. First, the Department of Health and Social Care's formal response to the SIT Committee report, and whether it commits to the February 2027 break clause, to a procurement of a UK-owned alternative, or to the status quo.[3] Second, the next ICB or trust to publicly refuse the FDP, and whether the Greater Manchester template (defer, gather public feedback, document the value gap, refuse on the published record) is replicated.[2] Third, the next published primary source on the Palantir "admin role" access, and whether NHS England publishes a public-facing summary of the data-access structure that closes the gap between the "direct care" framing and the actual access structure inside the FDP.[1] The answers to those three questions will determine whether the 2026 cycle is a one-off pressure spike or the standing pattern for the next decade of NHS data infrastructure.

Sources

1. The Register: "NHS patients can't opt out of Palantir's data platform, but their hospital can" by SA Mathieson, June 13, 2026 (primary source for Preet Kaur Gill's June 13 House of Commons answers, the National Data Opt-Out policy position, the trust-level opt-out confirmation, the 168-of-214 NHS trusts figure, the 40 written questions, the Palantir "admin role" change, the FT-confirmed briefing document on "risk of loss of public confidence," and the February 2027 contract expiry date)
2. The Register: "Greater Manchester still says no to NHS data platform with Palantir at its heart" by Lindsay Clark, May 13, 2026 (primary source for the 2.8 million population coverage, the 2025 ICB-level deferral, the FOI response on heightened public concern, and the in-house data-analytics-capability framing)
3. The Register: "UK lawmakers call on government to ditch Palantir NHS contract" by Lindsay Clark, June 3, 2026 (primary source for the House of Commons Science, Innovation and Technology Committee report, the recommendation to use the February 2027 break clause, the Dame Chi Onwurah chair statement, the Palantir/Thiel/Karp political-economy framing, the "UK-owned and UK-based providers" recommendation, and the 40 written questions figure)
4. openDemocracy: "Palantir: NHS data deal with US spy-tech firm puts patient trust at risk" by Rebecca Hilsenrath (acting editor) and others, November 21, 2023 (primary source for the original November 2023 opt-out asymmetry disclosure, the David Davis MP quote on anonymization, the Cori Crider / Foxglove framing, the Doctors' Association UK position via Dr David Nicholl, and the 2021 Foxglove + openDemocracy legal history)
5. Foxglove press release: "UK Govt must listen to MPs' call for end to Palantir contract" by Martha Dark (Co-Executive Director, Foxglove), June 3, 2026 (primary source for the cross-party SIT Committee framing, the 2027 break-clause recommendation, the "No Palantir In Our NHS" campaign, the Martha Dark quote on the urgent call to kick Palantir out, and the Foxglove legal-action history)
6. Medact: "Briefing: Concerns Regarding Palantir Technologies and NHS Data Systems," March 12, 2026 (primary source for the coalition briefing structure, the Good Law Project / Privacy International / Just Treatment / Corporate Watch / United Tech and Allied Workers Union / Amnesty International / Keep Our NHS Public endorsements, and the patient-trust-and-ethics framing)
7. Financial Times: "Palantir loses legal challenge against Swiss investigative magazine," June 12, 2026 (primary source for the Zurich Commercial Court dismissing the bulk of Palantir's claims against the magazine Republik over its data-handling and public-sector reporting, and for the framing of the case within Palantir's wider European pushback)
8. The Intercept: "NYC's Public Hospital System Won't Renew Its Palantir Contract" by Sam Biddle, March 24, 2026 (primary source for the New York City Health + Hospitals decision not to renew its roughly $4 million Palantir contract, the community-organizing and public-pressure campaign behind it, and the ICE/data-pipeline concerns that drove the review); see also American Friends Service Committee: "Facing public outrage, NYC hospital CEO announces Palantir contract will not be renewed" and Becker's Hospital Review: "NYC Health + Hospitals to end $4M Palantir contract"
9. SWI swissinfo.ch: "Palantir loses legal challenge against Swiss investigative magazine," June 2026 (source for the wider 2026 pattern of institutional and oversight pushback against Palantir across jurisdictions: Germany's armed forces excluding Palantir from contracts, Denmark and Netherlands officials seeking to reduce dependence, the UK NHS deal facing scrutiny, and the London mayor vetoing a Metropolitan Police contract)
10. NHS England, "Federated Data Platform FAQs: Can patients opt out of sharing their data with the Federated Data Platform?" (the standing public-facing policy source for the National Data Opt-Out position; the published FAQ text is the source of the "direct care" framing and the "secondary uses" carve-out the minister cited on June 13)

Related Coverage

• Palantir Just Lost a Swiss Court Fight Against a Magazine The June 13 SOS piece on Palantir's first international legal loss in a friendly jurisdiction. The NHS opt-out asymmetry, the SIT Committee report, and the Swiss court loss are the three largest Palantir pressure points of June 2026, and they all land in a 30-day window.
• NYC Kicks Palantir Out of Public Hospitals The March 2026 SOS piece on the NYC Health + Hospitals CEO's decision not to renew the $4 million Palantir contract. The community-organizing template that won the NYC hospital non-renewal is the same template the UK privacy campaign groups are now running against the FDP, and the SIT Committee report explicitly cites the same vendor-lock-in framing that drove the NYC reversal.
• Palantir NYC Hospitals Patient Data ICE Deportation The earlier SOS piece on the original NYC Health + Hospitals Palantir contract. The Palantir-ICE connection is the structural fact the SIT Committee report flagged, the fact the openDemocracy 2023 disclosure named, and the fact the trust-level opt-out is the working response to. The Greater Manchester ICB is the UK-side equivalent of the NYC Health + Hospitals community-board pressure.
• ICE Palantir ImmigrationOS Mission Creep Tracking Americans The US-side Palantir-ICE surveillance coverage. The SIT Committee report cites Palantir's ICE contract by name, the Foxglove press release cites the same contract, and the Medact briefing's Section 1 on "Allegations of complicity in human rights abuses" runs through ICE, border enforcement, and military operations. The opt-out asymmetry is the structural reason an NHS patient cannot use the National Data Opt-Out to keep their data out of a system whose US-side parent has been documented supporting those operations.
• Palantir Elite ICE Targeting App Confidence Scores The SOS piece on the ICE-targeting app deployment and the confidence-score framework. The confidence-score architecture is the same model the FDP's "admin role" inherits; the same logic that lets an analyst score a confidence band on an immigration target is the logic that lets an admin user decide whether to look at an identifiable patient record. The Medact briefing names this architectural pattern in Section 2 on procurement and government contracts.
• DHS Palantir Billion Dollar AI Surveillance Shopping Spree The US-side DHS procurement pattern. The SIT Committee report and the Foxglove press release both name Palantir's US-side federal contracts as the structural reason the UK government is being asked to reconsider its UK-side public-sector role. The opt-out asymmetry is the UK public's primary tool, but the procurement-pattern argument is what the international pressure on Palantir's overall public-sector footprint runs through.
• Amnesty International Palantir Babel Street Surveillance The Amnesty UK and Amnesty International coverage of Palantir's role in the FDP and Babel Street's parallel surveillance role. The Medact briefing's Section 1 cites Amnesty as the source for the "Allegations of complicity in human rights abuses" framing, and the opt-out asymmetry is the procedural gap Amnesty has been flagging since the contract was first awarded.