The UK’s ambition to become a sovereign AI powerhouse is sparking debate across government and industry about everything from data centers and cloud infrastructure to semiconductors, investment and regulation.
But at a recent OpenUK roundtable on digital sovereignty and AI (pictured), experts asked whether open source technology could help organizations avoid dependence on individual suppliers while maintaining greater control over the systems they rely on.
Although speakers did not agree on a single definition of sovereignty, many argued that transparency, interoperability, and resilience will become increasingly important as AI is integrated into public services and critical infrastructure.
definition of sovereignty
Amanda Block, Chief Executive of OpenUK, opened the discussion by highlighting the challenges posed by the widespread use of the term sovereignty.
“These conversations are often at cross purposes because we don’t have a good foundation on what sovereignty is,” she said.
Block argued that discussions about data sovereignty, digital sovereignty and AI sovereignty are often confused, even though they refer to different issues. At the same time, she cautioned against viewing sovereignty as something that can be achieved alone.
“One of the best things about creating value with open source is global collaboration,” she said. “That technology, which is absolutely necessary to become sovereign, would not exist for us today. Without global cooperation, it would not exist.”
Beyond geography
Several participants argued that sovereignty should not be reduced to where technology is hosted or where suppliers are headquartered.
Johnny Williams, Red Hat’s public sector principal digital adviser, said sovereignty should instead be viewed through the lens of control and flexibility.
“Fundamentally, I believe that digital sovereignty is really about agency, choice and control,” he said.
Williams argued that organizations need to understand the technology they rely on and maintain the ability to make informed decisions about their future direction.
He also warned against what he called “sovereignty washing,” where sovereignty claims are limited to geography and ignore broader dependencies.
“This is not a geography issue,” he said.
Instead, Williams argued, governments and organizations should focus on understanding where dependencies exist and ensuring they have meaningful choices about the technologies they use.
Sovereignty through resilience
Discussions also considered whether complete technological independence is a realistic goal.
Matthew Crawford, director of IP and AI compliance at British semiconductor company Arm, argued that the complexity of modern AI systems means it is unlikely that a single country will be able to control all the components.
“I don’t think it’s possible for one country to manage all the individual components of an AI model,” he said.
Mr Crawford suggested the government should focus on resilience rather than pursuing self-sufficiency.
“Digital sovereignty is about ensuring resilience. If any of these components go down, can we ensure that the countries and people using this technology can continue to use it?”
Why open source matters
For Duncan Currie, who leads partnerships and alliances at SUSE, sovereignty has become a mainstream concern among organizations deploying AI and digital services.
“All our conversations with carriers always come back to sovereignty, both in terms of data, technology and AI,” he said.
If you like this content…
Curry argued that much of the debate ultimately focuses on who controls the data and the technology that processes it.
“It comes down to who controls the data, who has the sovereign right to do it, and who the algorithmic models underpin it.”
He also emphasized the role open source can play in building trust and transparency.
“Open source is undeniable,” he said. “The way to build trust in code is not to put it behind a paywall.”
Britain’s innovation challenge
The discussion also touched on the UK’s ability to develop and commercialize new technologies. Matt Barker, founder and chief executive of Bolt MCP, argued that the UK already has the skills and expertise needed to play a key role in the AI ecosystem.
“We have 100 percent brain power and technical power,” he said.
Barker pointed to the Model Context Protocol (MCP), which is becoming an increasingly important standard for connecting AI systems to external tools and data sources.
“MCP was built by Austrians in London,” he said.
The challenge for Mr Barker is not a lack of innovation, but rather ensuring that successful ideas can be developed and scaled up in the UK.
Rethink the basics
Technology advisor and entrepreneur Sir Nat Way suggested that some of today’s debates about sovereignty stem from decisions made much earlier in the development of the internet itself.
“The Internet was built with an apocalyptic world in mind, where there are no states,” he says.
Wei argued that concepts such as identity, trust, and sovereignty are not fully baked into the Internet’s architecture.
As governments increasingly seek greater control over AI, data and digital infrastructure, they need to think more carefully about how these principles are reflected in future technologies, he suggested.
“We haven’t thought about this enough on a technical level,” he said.
Interdependence, not isolation
One of the conclusions of the discussion was that sovereignty should not be confused with technological isolation.
Participants generally agreed that international cooperation will continue to be essential for future innovations, especially in the field of AI.
The challenge, they argued, is to ensure sufficient control, transparency, and resilience for governments, businesses, and public sector organizations as they participate in the global technology ecosystem.
Mr Williams suggested that the UK should focus on influencing ecosystems rather than trying to separate them from them.
“We can make a net contribution to the global code,” he said.
