Executive Summary

• Kao Data’s Chief Business Officer, Clinton Hasell, shares his 2026 prediction on the data centre industry, dubbing 2026 the “year of reckoning.”
• He states there’s no doubt that the UK has innovation and ambition; it’s whether we have the infrastructure, regulatory environment, and political willpower to turn that ambition into reality.
• While the media focuses on the energy and sustainability bottlenecks, the skills gap will emerge as a constraint of the same magnitude next year.

Kao Data’s Chief Business Officer, Clinton Hasell, has shared his 2026 predictions with Data Centre Insight editor, Lauren Raybould, as we look ahead to the new year and the changes it will bring to the data centre industry.

2025 has been a year of bold ambitions for UK AI infrastructure. Matt Clifford’s AI Opportunities Action Plan captured the imagination of an industry hungry for direction, promising to transform Britain into an AI superpower and capture a share of the £400 billion international AI prize. AI Growth Zones were announced, ministerial visits to data centres made headlines, and the rhetoric around sovereign compute has never been stronger.

2026 Predictions

But as we look ahead to 2026, the question isn’t whether the UK has ambition – it’s whether we have the infrastructure, regulatory environment, and political willpower to turn that ambition into reality. From my vantage point, having spent over three decades in the data centre and digital infrastructure sector, I believe 2026 will be the year of reckoning: the moment when Britain must decide if it’s serious about being an AI maker, or whether we’re content to become an AI taker – dependent on compute capacity built, owned and operated outside the UK.

Energy Will Become the Critical Battleground

The single greatest determinant of the UK’s AI future isn’t compute power or talent, it’s energy. In 2026, the industry will be forced to confront a fundamental truth: while we debate policy frameworks, our competitors are building.

The numbers are stark. The UK currently pays approximately £1.8 billion annually per gigawatt of power – four times more than the United States and the Nordic countries. Our electricity market price is set by gas 98% of the time, despite 41% of our generation now coming from renewables. This pricing anomaly isn’t just an inconvenience; it’s an existential threat to our nation’s AI ambitions.

I predict that 2026 will mark the point at which this issue reaches a critical point. Data centre operators and hyperscalers will increasingly vote with their feet, directing frontier AI workloads, and the associated investment, to markets with more competitive energy pricing. We’ve already seen UK-headquartered companies place significant compute capacity in Norway and other Nordic markets. Without urgent reform to include data centres within Energy Intensive Industries eligibility criteria and address the fundamental disconnect between renewable generation and consumer pricing, this exodus will accelerate.

AI Growth Zones: From Announcement to Execution

The government’s AI Growth Zone programme has generated genuine excitement within the industry, with over 100+ applications from sites across the country. In 2026, we’ll see whether this enthusiasm translates into operational infrastructure or remains a policy framework in search of delivery.

The critical success factor will be power. AI Growth Zones must offer meaningful energy cost advantages and accelerated grid connections to attract the industrial-scale investment required. The government’s recent announcements around targeted pricing support mechanisms and allowing developers to build their own high-voltage infrastructure are promising, but implementation will determine success.

I expect 2026 to see at least two or three AI Growth Zones move from designation to active development, with the Northeast likely to lead. However, the actual test will be whether these zones can secure anchor tenants – the hyperscalers and neoclouds whose commitment provides clear signals to the market that the UK is genuinely competitive for frontier AI workloads.

The Grid Constraint Crunch

Grid connection delays of five to fifteen years in key metropolitan areas represent a fundamental barrier to data centre development. In 2026, this constraint will force the industry to adopt alternative power solutions.

I predict we’ll see significant growth in behind-the-meter generation, co-location with renewable energy sources, and direct private wire arrangements. At Kao Data, our partnership with Downing to develop a 40MW solar farm adjacent to our Harlow campus exemplifies this approach – taking proactive action rather than waiting for grid infrastructure that may never arrive.

The government’s commitment to reforming the connections process and allowing strategic projects to build their own high-voltage infrastructure could be transformational, but 2026 will test whether regulatory reform can keep pace with industry demand.

Skills: The Hidden Constraint

While energy dominates the headlines, the UK’s data centre skills shortage will emerge as an equally pressing constraint in 2026. With engineering and construction labour markets already tight, and a domestic pipeline of AI-ready facilities that demands specialist expertise, we don’t have enough qualified people to build and operate the infrastructure Britain needs.

Expect 2026 to see a fundamental shift in how the industry approaches talent development. Apprenticeship programmes will expand significantly, and the sector will need to dramatically improve its diversity – with women currently representing less than 8% of the workforce, we’re effectively operating with one hand tied behind our back.

Industry-led training initiatives, partnerships with further education providers, and targeted programmes to attract talent from adjacent sectors will become strategic priorities rather than nice-to-have, responsible business activities.

Sustainability: Integrated, Not Optional

The sustainability conversation in data centres will mature significantly in 2026. Operators who treat environmental performance as a compliance exercise will find themselves increasingly uncompetitive, as customers, particularly those in regulated industries, demand verified renewable energy sourcing and water efficiency.

Liquid cooling will transition from emerging technology to standard specification for facilities engineered for AI. The thermal demands of high-density GPU deployments cannot be met with traditional air-cooling approaches, and operators who have invested ahead of this curve will hold a significant competitive advantage.

The integration of on-site renewable generation, battery energy storage systems, and sophisticated demand management will become baseline expectations rather than differentiators. In 2026, the question won’t be whether a data centre is sustainable, but how comprehensively its environmental credentials can be verified and audited.

The Year of Decision

Britain possesses remarkable advantages in the global AI race: world-class research institutions, a thriving AI startup ecosystem, deep pools of technical talent, and a regulatory environment that has historically balanced innovation with responsibility.

But advantages only matter if you leverage them. In 2026, the UK will either demonstrate that it can translate policy ambition into physical infrastructure, addressing the energy pricing anomalies, grid constraints and skills gaps that currently impede development, or we will watch the AI future being built elsewhere.

The choice is ours. The clock is ticking. 2026 is the year we must decide: are we building AI, or just simply using it?

Clinton Hasell is Chief Business Officer at Kao Data, the specialist developer and operator of data centres engineered for AI and advanced computing. With over 30 years’ experience in the data centre, telecoms and digital infrastructure sectors, Clinton leads UK and European site development and core business operations.