Palantir's 'Glowing' Claims Aren't Evidence

In the world of health tech, there is a wide chasm between a vendor's marketing pitch and clinical evidence. Palantir is currently attempting to bridge that gap with rhetoric rather than rigor, and NHS England is right to push back.

As first reported by The Register, Palantir UK senior vice president Tom Watson has been making public claims that the company's Federated Data Platform (FDP) is driving tangible changes and benefits across the health service. Watson has pointed to figures suggesting 137 trusts are reporting benefits and highlighted specific examples from Chelsea and Westminster Hospital, University Hospitals of Derby and Burton, and Northampton General Hospital.

But here is the problem: these are self-reported wins. In a video cited by The Register, Watson admitted that Palantir cannot prove every procedure happened because of their software. While he argues we shouldn't ignore what hospitals say, in a high-stakes healthcare environment, anecdotal success is not the same as causal efficacy.

We have already seen the friction between Palantir's narrative and the actual data. The Register reports that while NHS England claimed an Inpatient Care Coordination Solution tool helped hospitals perform 111,589 additional procedures, a Freedom of Information request revealed that nearly a third of trusts using the platform were actually performing fewer procedures than before the system went live.

Even the Office for Statistics Regulation (OSR) has stepped in. As reported by The Register, the OSR noted that NHS England had to admit it needed to communicate FDP performance claims more clearly. Crucially, NHS England published a caveat on its own website stating it could not draw conclusions about cause and effect because other variables—such as industrial action or staffing levels—had not been controlled for.

It is telling that Tom Watson, in a statement on X, described the OSR's lack of a controlled study as a "fair note of caution," yet he continued to repeat benefit claims without mentioning those same caveats. This is classic vendor-driven hype: highlighting the wins while burying the methodology.

NHS England CEO Sir Jim Mackey is now planning to tell Palantir to clear any future benefit claims with his team. While Mackey admitted to the Health and Social Care Committee that this may be contractually difficult to enforce, the move is a necessary safeguard. The £330 million contract awarded to the Palantir-led consortium in November 2023 is too significant to be justified by press releases.

The only path forward is independent, peer-reviewed validation. I am encouraged to see that NHS England awarded a contract to Imperial College Projects in February 2026 to independently evaluate the FDP. However, The Register notes that these findings aren't expected until after ministers decide whether to extend the contract, which expires in March 2027.

Scaling a tool based on the word of the company selling it is a recipe for failure. If the FDP is truly the efficiency engine Palantir claims it to be, it should be able to withstand the scrutiny of an independent audit. Until then, the NHS should treat Palantir's "glowing" assessments as exactly what they are: marketing.