Gartner warns of rising lock-in to regional AI stacks

Gartner forecasts that more than a third of countries will become locked into region-specific AI platforms within the next two years, as governments and large organisations respond to geopolitical, regulatory and security pressures.

The analyst group said 35% of countries will be locked into AI platforms that rely on proprietary contextual data by 2027. It also predicted platform lock-in will rise to 35% from 5% over the same period.

Gartner linked the shift to increased focus on digital sovereignty and a move towards domestic AI infrastructure. It cited a desire for alternatives to approaches it characterised as closed and led by the United States.

"Countries with digital sovereignty goals are increasing investment in domestic AI stacks as they look for alternatives to the closed U.S. model, including computing power, data centres, infrastructure and models aligned with local laws, culture and region," said Gaurav Gupta, VP Analyst, Gartner.

Regional platforms

Gartner said decision makers increasingly judge AI platforms on "trust and cultural fit". It said these factors now compete with scale and breadth of training data when organisations select AI systems for public services and regulated sectors.

"Trust and cultural fit are emerging as key criteria. Decision makers are prioritising AI platforms that align with local values, regulatory frameworks, and user expectations over those with the largest training datasets," said Gupta.

The firm argued that localised models can outperform global models in specific use cases. It highlighted education, legal compliance and public services, particularly for non-English languages. It said regional large language models can deliver more contextual value in those settings.

In Gartner's view, the resulting market structure favours region-specific platforms that blend models, data and infrastructure tuned to local requirements. It also increases the risk that organisations become tied to a particular ecosystem. That can affect switching costs and procurement flexibility.

Spending targets

Gartner also set an expectation for government investment linked to AI sovereignty. It predicted that nations establishing a sovereign AI stack will need to spend at least 1% of their GDP on AI infrastructure by 2029.

It described AI sovereignty as the ability of a nation or organisation to independently control how AI is developed, deployed and used within geographical boundaries. Gartner said sovereign AI efforts now extend beyond policy. They include local compute, data storage, and domestic or regionally aligned model development.

The analyst group cited a range of drivers. These included regulatory pressure, geopolitics, cloud localisation requirements, national AI missions, corporate risk concerns and national security considerations. Gartner also pointed to competitive dynamics. It said fear of falling behind in an AI race would push faster investment and attempts at self-sufficiency across the AI stack.

Gartner said the shift could reduce international collaboration. It also said it could create duplication of effort as countries and blocs build parallel systems and supply chains. It linked this to concerns among non-Western customers about alignment and what it described as overly Western influence.

Infrastructure build

Gartner placed data centres at the centre of its outlook for sovereign AI. It said sovereign AI strategies require domestic capacity for training and running models, and local infrastructure for sensitive data.

"Data centres and AI factory infrastructure form the critical backbone of the AI stack that enables AI sovereignty, " said Gupta. "As a result, data centres and AI factory infrastructure will see explosive build-up and investment going forward, propelling a few companies that control the AI stack to achieve double-digit, trillion dollar valuations."

The comments reflect a broader push by governments to map critical technologies to local industrial strategy. In AI, that often includes procurement rules, data residency requirements, and an emphasis on national cloud and data centre capacity.

CIO guidance

Gartner's note included guidance for CIOs and senior technology leaders operating across multiple jurisdictions. It recommended "model agnostic workflows" and orchestration layers that allow switching between large language models across regions and vendors.

It also urged leaders to ensure AI governance, data residency and model tuning practices meet country-specific legal, cultural and linguistic requirements. The firm recommended relationships with national cloud providers, regional model vendors and sovereign AI stack suppliers in priority markets, alongside a vetted partner list.

Gartner also advised organisations to monitor AI legislation, data sovereignty rules and emerging standards that could affect where models can run and how user data is processed. The firm expects these requirements to fragment deployments across markets and shape vendor selection in the next few years.

Gartner said its analysts will discuss data and analytics trends at its Data & Analytics Summits during 2026.