BT on why it’s betting big on keeping AI on British soil

British telecoms giant BT is making a major push into sovereign AI infrastructure, with plans to keep data, models, and operations firmly on UK soil.

In responses to HotMinute, a BT spokesperson outlined the company’s strategy for building ‘sovereign AI capability’ designed so that infrastructure, data, security governance, and operational control all remain in the UK, under UK law and jurisdiction.

The initiative includes the construction of 14 megawatts of AI compute capacity across three strategic BT sites in the United Kingdom.

These facilities will be accessed via BT’s secure national network and governed through the company’s existing UK security, compliance, and operational frameworks. Services will be managed by UK-based, security-cleared teams.

As a result, customer data and AI workloads will remain subject to UK legal and regulatory processes.

Why sovereign AI matters

BT highlighted the increasing importance of digital sovereignty for businesses as AI becomes central to operations.

Organisations want greater oversight of where their data is processed and how services are managed, particularly in regulated sectors handling sensitive information.

“Digital sovereignty is becoming increasingly important to businesses as technology becomes more deeply embedded in how organisations operate,” the spokesperson said. “They want clearer oversight of how and where their data, platforms, and services are managed.”

Running AI models in the UK, rather than relying on foreign data centres, reduces legal uncertainty, compliance risks, and enhances resilience, according to BT. This is especially relevant as global regulations around data privacy and AI governance continue to evolve.

Industrial-scale capability

The 14 MW deployment represents serious computing muscle. BT described it as “industrial-scale AI, enough to run many large models, serve thousands of users continuously, and embed AI into real-world services, not just pilots.”

Built on Rackspace Technology’s UK data centre infrastructure, it delivers compute, storage and backup capabilities with UK-based, security-cleared teams handling migration, operations, and compliance.

While specific details on GPU types (other than it being Nvidia hardware) or exact processing power were not disclosed, the scale positions BT as a significant player in the UK’s domestic AI infrastructure push.

BT’s move aligns with broader UK government efforts to build domestic AI capability.

The government’s AI Opportunities Action Plan has pushed for investing in sovereign data centre infrastructure to drive economic growth, resilience, and national security.

The UK has already committed £500 million through its Sovereign AI fund, recognising AI’s potential to boost the economy and create jobs. BT said its sovereign portfolio aligns directly with these national aims.

The announcement comes as governments worldwide grow wary of depending on hyperscalers headquartered in other jurisdictions, especially amid geopolitical tensions and varying data protection standards.

By keeping AI workloads under UK control, BT aims to offer enterprises, particularly in finance, healthcare, and public sectors, the transformative benefits of AI without the associated sovereignty risks.

The launch coincides with new research from Assembly Research, commissioned by BT, which highlights the significant economic potential of digital sovereignty.

The report estimates that addressing concerns over data security could unlock £18 billion in productivity benefits for the UK by accelerating secure AI adoption.

It also points to broader commercial opportunities: accelerated investment in UK-based data centres could generate £14.6 billion by 2030, while expansion of sovereign cloud services could add another £13.6 billion over the next five years.

Now read: Major new Internet cable planned between the UK and Iceland