Digital and Technology Policy: National Sovereignty

Baroness Stowell of Beeston Excerpts

Thursday 23rd July 2026 Lords Chamber

[Baroness Stowell of Beeston (Con)]

My Lords, I am grateful to the noble Baroness, Lady Kidron, for facilitating this debate on such an important topic. It is, as always, a pleasure to follow my noble friend, who makes such thoughtful contributions.

I want to concentrate on artificial intelligence. As I think we all know, alongside magnificent opportunities, it presents many risks. The greatest of those risks is not intelligence or what the technology can do, but the concentration of power in too few companies, in too few countries and, ultimately, in too few hands. We have seen the consequences—the noble Baroness, Lady Kidron, has talked about some of them—when digital markets become dominated by a small number of powerful platforms: reduced competition, weaker innovation and increased dependency.

The question, therefore, is not whether concentration of power matters—it plainly does—but how we respond. I share with many noble Lords the concern that the UK should not become dependent on a handful of foreign tech companies for critical digital capabilities, but I am not convinced that the answer is more regulation. The largest and most dominant frontier model developers benefit if Governments believe that only they can be allowed to develop safely the most powerful AI. They benefit from a resulting policy debate in which ever greater risks are assumed to require ever greater regulation. That, to me, is the real danger of regulatory capture. Regulation introduced in the name of safety can become a barrier to entry that protects today’s dominant platforms from tomorrow’s competitors.

Before we legislate, we should ask not only whether a risk is sufficiently real and significant to justify intervention—of course, this week’s OpenAI breakout cyber attack makes that question a live one—but whether the proposed intervention will strengthen competition or simply reinforce existing market power. The answer to concentration is competition. The open-weight foundation models are the greatest enabler of competition that AI has yet produced. They fundamentally change the economics of innovation by lowering the cost of deployment and entry—I refer noble Lords to an article in the Washington Post this week by Bill Gurley. These open models allow many more companies, including British start-ups, to build world-class AI applications without first having to spend billions training frontier models. Recent developments have demonstrated that this is no longer theoretical. The emergence of highly capable, low-cost open- foundation models, including from China, shows that cutting-edge AI is no longer confined to a handful of closed proprietary systems. We should welcome that, not because we wish to depend on China—we obviously do not—but because competition is the surest antidote to concentration of power.

Britain does not need to own a frontier model to achieve AI sovereignty. We need the freedom to choose between competing models, the ability to build on them and the capability to deploy them securely, on our terms, into our economy and public services. As I and many other noble Lords have argued before, Britain’s challenge has never been inventing technology; it has been scaling technology companies. Open-foundation models give more British firms the opportunity to compete, grow and retain strategic capability here, as in not bought or moved elsewhere. That is why competition policy matters and why, only where relevant and necessary, the CMA must use its powers in the Digital Markets, Competition and Consumers Act to ensure that AI markets remain open, competitive and capable of supporting new entrants.

Competition must of course go hand in hand with protection for creators, and the arrival of open foundation models makes getting copyright right more, not less, important. Creators should have the technological means to know when their work is being accessed, to decide whether and at what terms it may be used for AI training, and to receive fair compensation when it is. Measures such as the Private Member’s Bill that my right honourable friend Damian Hinds is introducing in another place about online software access and transparency point towards the kind of targeted transparency that can help achieve those objectives.

The UK has an opportunity to take a distinctive path, embracing open foundation models, protecting intellectual property, using our competition powers robustly and creating the best environment for AI companies to start, scale and compete. If sovereignty is our objective, competition, not concentration, must be our strategy, and competition in AI will depend on whether we embrace the technology that enables it, which will be open-weight foundation models.