UK firms back open source to bolster AI sovereignty

Red Hat has published survey findings on how UK organisations are approaching AI sovereignty, pointing to a gap between planning for provider disruption and maintaining business continuity.

Research among 100 UK IT decision makers found that 67% have a defined exit strategy if their main AI provider restricts access to services. Even so, 43% said such a change would still have a moderate to significant impact on business continuity.

The figures suggest many businesses have contingency plans on paper but remain reliant on a narrow set of AI suppliers. Red Hat framed the issue around sovereignty: control over data, infrastructure and provider relationships as AI becomes more embedded in business operations.

Governance gap

The study also found that agentic AI is already in wide use in the UK, with 87% of respondents reporting use of agentic AI systems. Red Hat described these as tools that can take actions and trigger workflows autonomously.

Governance appears to lag behind adoption. Only 25% of UK respondents said they have strong governance frameworks for agentic AI. Another 43% said they have some governance with gaps, while 17% described it as basic or minimal.

This suggests a sizeable share of organisations are deploying systems with some degree of autonomy without robust oversight. Across the five European markets covered by the survey, 64% of organisations reported having either some governance or strong governance in place.

Data visibility

On data and infrastructure, 93% of UK respondents said they have complete or partial visibility over where data is stored, processed and potentially accessible. However, for 45% that visibility is only partial, while 48% reported complete visibility.

Compared with other markets, the UK ranked below Germany on this measure, where 97% reported complete or partial visibility. The Netherlands and Italy both recorded 90%.

These findings come as businesses face increasing scrutiny over where data resides, who can access it and how AI systems are governed. For many companies, the question is no longer whether to use AI, but how to do so without becoming overly dependent on a single platform provider or losing sight of data handling across multiple environments.

Open source support

The survey found strong support in the UK for open source approaches to AI. Four in five respondents said open source offers greater control over how AI is built and where it runs.

Over the next three years, respondents said the most valuable benefits of open source for building trust in AI would be transparency and easier auditability, cited by 87%, followed by greater customisation for business and regulatory needs at 82%, and greater control over how AI is built and where it runs at 80%.

Support for regulation was also pronounced. Some 89% of UK respondents said public policy and regulation should mandate open source principles such as transparency, auditability and open source licensing to support AI sovereignty.

That was above the EMEA average of 77% and ahead of France at 70% and Germany at 72%. The figures indicate relatively strong appetite among UK technology leaders for rules that set standards around openness and accountability in AI systems.

Joanna Hodgson, Country Manager, UK, Red Hat, commented on the findings.

"AI platforms are increasingly part of UK organisations' critical infrastructure. Many have written exit strategies in preparation for any challenges, but our survey shows that actually executing a switch without disruption remains difficult. To close that gap, enterprises need greater control over how and where AI runs, and a consistent way to govern fast-moving technologies like agentic AI. Enterprise open source gives UK businesses the transparency, flexibility and shared innovation they need to treat AI as a resilient, sovereign capability," said Joanna Hodgson, Country Manager, UK, Red Hat.

The wider regional picture suggests similar concerns are being discussed across European businesses as AI moves from pilot projects into operational use. The research covered the UK, France, Germany, Italy and the Netherlands, with 500 IT decision makers surveyed in total.

Hans Roth, Senior Vice President & General Manager EMEA, Red Hat, said the debate has moved beyond early experimentation.

"Across EMEA, boardroom conversation has moved beyond experimentation to how AI can be deployed in a way that meets sovereignty, security and regulatory expectations. The survey results show strong support for open source principles and for clear policy frameworks that embed transparency and auditability into AI. That tells us organisations are not looking for another closed, one-size-fits-all stack; they want the freedom to combine different models, accelerators and clouds while staying in control," said Roth.