King’s Speech

Lord Davies of Brixton Excerpts

Wednesday 20th May 2026 Lords Chamber

[Lord Davies of Brixton (Lab)]

It is a pleasure and an honour to take part in this interesting, if wide-ranging, debate. I will focus my remarks on the issue that was raised by my noble friend the Minister in opening: namely, the single patient record proposal, which will be part of the NHS modernisation Bill. I have the NHS app on my phone, as well as the patient access app from my local practice, and I have my chart from Guy’s and St Thomas’. I wish they would talk to each other. I am also a participant in UK Biobank, so that has my data as well.

While I welcome the proposal for the single patient record, I do not believe that Palantir is a suitable place to secure our health information and data. In truth, I am probably a bit late, because Palantir is already involved through the NHS federated data platform. However, its involvement is tightly constrained. What is being proposed here—if the data were handed over to Palantir—would be a very different situation, which raises not just technical issues but constitutional, ethical and democratic matters.

First, we need to understand the nature of the organisation. It was founded on the single premise that vast quantities of data, properly analysed, confer decisive commercial advantages. Historically, its customers have been the CIA, the NSA in the States and the US Department of Defense. Its software has been used in ways which, in my view, fall outside UK norms. This is not incidental to the company: it is at the heart of what that company does.

The second issue of concern is that the commercial incentives are deeply troubling. It is important to understand that it is not just the data that they benefit from: it is the links, the structure and the way in which the system works. That is where the real money is to be made. The data is just a means of moving towards a knowledge of the system on which we are based.

The third problem, which is well known, is the issue of dependency. We have seen this all too often. Once you adopt a particular software approach, you are very heavily locked in. If Palantir were to be involved, we would be locked in. It is not simple to walk away. To me, this makes it clear that we need an in-house solution. It is not beyond the capacity of an organisation the size of the NHS to develop its own software, in-house, protecting national interests.

Fourthly, there are questions about the governance framework. A particular concern is whether US surveillance law compels Palantir, as a foreign organisation, to disclose NHS data to American intelligence agencies. I think we need a clear and unambiguous position on that.

Finally, there is an argument of principle. Health data is not a resource to be managed by whoever submits the most competitive tender. It is generated in moments of profound vulnerability: illness, birth, death and mental crisis. The relationship between the patient and the health system rests on a compact of trust; the involvement of Palantir risks severely breaking that trust.