Digital and Technology Policy: National Sovereignty

Lord Vaizey of Didcot Excerpts

Thursday 23rd July 2026 Lords Chamber

[Lord Vaizey of Didcot (Con)]

My Lords, I am grateful for the opportunity to take part in this debate, and I add to the plaudits directed at the noble Baroness, Lady Kidron, who has prosecuted her agenda on digital safety and digital rights for more than a decade with acute skill and has certainly moved the dial on many occasions. May I also say how pleased I am that the Minister has survived the reshuffle cull? As she knows, I had a small desire that she would move to the environment department, so that we could refer to her as Baroness Lloyd of Effra of Defra, but we will have to wait. And may I say how sad I am that the noble Baroness, Lady Chapman, has left the Government? She was an absolutely excellent Minister. I declare my interest as an adviser to Digital Futures, which deploys AI engineers into companies and assesses their AI capabilities.

In November, the House of Lords Science and Tech Committee described the UK tech economy as an incubator economy. More recently, Amanda Brock from OpenUK made the point that we should not seek, when we talk about AI sovereignty, to build the whole AI tech stack by ourselves. Somewhere between those two statements lies the way forward, because we are in an interconnected world: we cannot simply ignore the rest of the world, but we must do better than be the incubator economy we have been for so many years. We must support British companies to scale and grow in this area.

Let me make five recommendations to the Government. The first is obviously to invest in our infrastructure: to invest in the UK companies that are making the chips, building the data centres and investing in the cloud. I was lucky recently to meet a British company, Fractile, which makes chips. It is a British company, based between Bristol, London and Cambridge, and it has recently raised $200 million, but it plans to stay and grow in Britain. This is exactly the kind of company that we should be supporting. As for not building the stack ourselves, let us look at our competitive advantage, of which Fractile is a good example. It is in a very niche business and has identified a need for a particular chip to support large language models which uses less energy and generates less heat. It is this kind of small infrastructure play that can reap huge dividends for the UK AI economy and is exactly what we should be looking at. We should be looking at grid connections and at building British data centres.

I also echo the earlier comment about creating a national data sovereignty platform. It is absolutely right that one of our unique advantages in this space is the access we have to the data of British citizens, which should be handled carefully but which gives us an enormous advantage, particularly in areas such as the NHS.

The third way forward is to invest as much as possible. I am delighted that the Government have set up sovereign AI—it is only £500 million, but that is still a significant sum; it is less than the €5 billion that the EU has set aside, but we are just a small island—and the fund that has been set aside for AI hardware. I hope that the Minister will reassure us that that £500 million will be fully allocated to sovereign AI and clear up some of the confusion. Her colleague said yesterday that it was moving to DBIST, but I am now told that it is moving to the Cabinet Office. Some clarity on that would be useful.

I referred earlier to my work with Digital Futures. Skills are hugely important in this area. I gather that there are predicted to be around 900,000 vacancies in the tech economy in the next few years and there was a drop in engineering apprentices of some 42% over the last few years. A focus on building the skill economy that we so vitally need to support these industries is important.

Finally, my killer point is that the best way forward would be to abolish the government department that is laser-focused on science, innovation and technology, and to scatter its constituent parts across three different departments and No. 10. That would be a masterstroke, and would allow our civil servants to spend six months trying to sort out where on earth they are meant to be going, rather than implementing policy and delivering. The Minister will not be surprised that I have descended into my default state of facetiousness. The abolition of DSIT is a huge error, considering the urgency this agenda requires and that it has taken three years to get this department up and running and to find its feet. It was a beacon of excellence in Whitehall; many people in the tech economy are sad to see its demise.