Interest in sovereign cloud services has followed a bumpy trajectory over the past decade, both in the public sector and private enterprises.
In the early days of cloud computing, fear, uncertainty and doubt reigned supreme in boardrooms and IT departments, as CIOs and IT decision-makers grappled with the notion of migrating workloads off physical hardware, housed in privately owned datacentres, to multi-tenanted public clouds, operated by overseas tech firms.
The situation fuelled interest in private cloud deployments, as they allowed enterprise IT departments to tap into the benefits of public cloud while maintaining a greater degree of control over the infrastructure and environment hosting their workloads.
Around this time, the domestic cloud markets in many countries, including the UK, began to flourish as providers emerged that specialised in the provision of locally hosted, sovereign public cloud services.
These setups provided UK-based enterprise IT buyers, for example, with assurances that their applications and workloads would be hosted in UK-based datacentres, whose operations would be governed by UK laws and regulations, with no access permitted by overseas governments or entities.
In time, enterprises became more at ease with using cloud technologies offered by US tech giants, such as Amazon Web Services, Microsoft and Salesforce.
However, a degree of trepidation regarding their usage persisted until these firms opened their own UK datacentres between 2014 and 2016, which offered users access to locally hosted versions of their services.
As documented by Computer Weekly, this development prompted a downturn in demand for sovereign cloud services from domestic UK cloud providers, which in turn led to some going out of business and others being acquired, or having to pivot their business models accordingly.
Growing concerns over sovereignty
Over the years, the hyperscale cloud providers have continued to see demand for their UK-hosted services grow, despite rising concerns about how much of the UK’s data is now being hosted in cloud environments operated by overseas cloud firms.
Such concerns have been heightened more recently by revelations, such as those uncovered by Computer Weekly, about Microsoft’s inability to guarantee the sovereignty of policing and other forms of public sector data stored within its Azure public cloud platform.
Geopolitical concerns are also focusing the minds of CIOs and IT decision-makers on making sure the cloud platforms they entrust with their data will keep it in-country and protected by domestic laws and regulations, says Oliver Hessel, head of big data products and cloud at hosting provider Ionos SE.
“Trade tensions between the US and China and policy shifts following US elections are creating supply chain uncertainties, particularly for AI hardware,” Hessel tells Computer Weekly.
“European businesses are responding with multicloud and diversification strategies, reducing reliance on non-European providers while mitigating risk, [and] digital sovereignty is increasingly tied to resilience. We’re also seeing cyber security concerns reinforce this trend,” he adds.
Against this backdrop, Hessel says hosting critical workloads locally and partnering with sovereign providers has “become a strategic safeguard” for a lot of enterprises already.
And forecast data from IT analyst house IDC, shared in its FutureScape 2026 research, suggests this trend is set to persist for some years to come.
For example, its data states that by 2028, 60% of organisations with digital sovereignty requirements will have migrated sensitive workloads to new cloud environments to reduce risk and increase autonomy.
IT market responds to rising digital sovereignty demands
The hyperscale cloud giants have responded to this trend by announcing and rolling out sovereign cloud regions, capable of providing users with access to versions of their services whose operations are governed by local laws and regulations.
Meanwhile, over the course of 2025, a new category of cloud company emerged, focused on the provision of more niche, sovereign cloud workloads.
The scaling of AI inside enterprises is becoming a major driver of sovereignty. As organisations become more ambitious, many will find that their existing infrastructure architectures cannot provide the assurances required for sensitive or regulated workloads Kate Hanaghan, TechMarketView
Known as neocloud providers, major players within this emerging category of tech suppliers include the likes of Nscale, CoreWeave and Carbon3ai, to name a few.
What these firms have in common is a focus on providing sovereign infrastructure that can be used to host high-performance computing (HPC) and artificial intelligence (AI) workloads.
According to IT market watcher Forrester, 2026 is set to be the year governments start to refine their AI strategies, with the company predicting that many will take a “tech nationalism” stance when selecting prospective AI suppliers. This suggests that many of these neocloud providers will be well-positioned to tap into this trend.
“Global digital norms will give way to tech nationalism when it comes to AI models,” Sam Higgins, vice-president and principal analyst at Forrester, writes in a blog post, outlining his tech predictions for 2026. “Amid geoeconomic fractures and AI disruption, 2026 is the year governments choose domestic-first, from model selection to hosting, rewriting AI procurement and compliance in the process.”
There are also various pieces of legislation coming down the line, including the EU AI Act, that all signal a growing preference for sovereign AI services, he continues.
“We [Forrester] expect that half of the G20 will mandate domestically tuned AI models for public sector services,” says Higgins. “For firms offering solutions to the public sector, prepare for this inevitability by undertaking a thorough inventory of your model provenance along with in-market hosting options for inferencing, or risk exclusion from future public sector opportunities.”
A similar set of preferences for sovereign AI services is also expected to play out in the private sector in 2026, says Kate Hanaghan, chief research officer at IT analyst house TechMarketView, as enterprises come to realise their existing infrastructure may no longer meet their needs.
“The scaling of AI inside enterprises is becoming a major driver of sovereignty,” she says. “As organisations become more ambitious, many will find that their existing infrastructure architectures cannot provide the assurances required for sensitive or regulated workloads.”
She continues: “As a result, concerns are shifting beyond data location and into the operational and digital layers, where the focus must move to governing the people, processes and technology components – such as the AI model, data pipelines and underlying infrastructure – that influence how the system behaves.”
As is the case in the public sector, regulatory pressure looks set to accelerate the adoption of sovereign AI services too, she adds.
“In financial services, there are concerns about a concentration on a small number of cloud providers. Meanwhile, operators of critical national infrastructure (CNI) will face rising scrutiny around resilience and data integrity. We expect this to reinforce the role of local assurance, including the appropriate use of UK-based hosting providers,” Hanaghan adds.
Speaking to Computer Weekly, Bjorn Hovland, chief operating officer at AI infrastructure company CIQ, says that in addition to regulatory and geopolitical pressures, there are also economic reasons why governments and enterprises will show a preference for sovereign AI in 2026.
“If a country outsources its ability to manage its own computing infrastructure, a foreign country or company could easily shut down key capabilities and infrastructure in the future,” warns Hovland.
If AI is the future of economic growth, countries that cannot run their own AI workloads risk being left behind and losing jobs/growth to more technologically advanced peers Bjorn Hovland, CIQ
“There is also a clear economic advantage. If AI is the future of economic growth, countries that cannot run their own AI workloads risk being left behind and losing jobs/growth to more technologically advanced peers.”
He continues: “As data continues to grow in value, governments are recognising that an inability to secure, manage and use the data generated by their countries is a major strategic vulnerability.”
Neocloud providers poised to respond
The UK government’s championing of UK-headquartered sovereign AI infrastructure provider Nscale, following the release of its January 2025 AI opportunities action plan policy paper, makes sense against such a backdrop.
Speaking to Computer Weekly, the company’s senior vice-president, Imran Shafi, says that while the analyst community is predicting an uptick in demand for sovereign AI services this year, work is already underway across Europe to meet this demand in 2026.
“Sovereign AI is increasingly about real-world implementation, rather than theory, and countries across Europe are moving beyond strategy documents and starting to build the physical infrastructure they’ll need to run critical AI systems on their own terms,” says Shafi.
“That means greater demand for AI datacentres, more clarity around operational sovereignty, and deeper collaboration with neighbours.”
Even so, Shafi is of the view that not every country will have the means, need or desire to build out their own sovereign AI infrastructure, which could see enterprises using “federated” setups to plug any gaps.
In such arrangements, AI models are trained collaboratively across multiple, decentralised, geographical locations, without the raw data itself needing to be migrated.
“This ensures sensitive workloads stay local, while energy-intensive requirements shift to regions like the Nordics or Iberia … [and this means] with the right investment in infrastructure and coordinated policy-making, Europe can stay competitive without compromising on values like privacy, sustainability and transparency.”
It is the view of INQ’s Hovland, though, that it won’t just be the neocloud providers that benefit most from enterprises and governments showing a preference for AI sovereign services in 2026.
“The two biggest winners of this trend are companies that can pivot to sovereign solutions quickly, especially hyperscalers, and companies that are building sovereign solutions by design,” he says.
“However, these are often complex technical solutions, so the winners will be enterprises that have anticipated this trend and have been building these solutions for some time.”
Sign Up For Daily Newsletter
Be keep up! Get the latest breaking news delivered straight to your inbox.
By signing up, you agree to our Terms of Use and acknowledge the data practices in our Privacy Policy. You may unsubscribe at any time.
