Research from IT infrastructure specialists Onnec, has found that 69% of data centre operators expect geopolitical instability and US tariffs to inflate the cost of building AI-ready data centres and lengthen timelines. A further 79% believe that the shortage of AI skills will delay data centre projects.

The study, based on a survey of 250 senior decision-makers from the UK, Ireland and the Nordics, warns the explosive rise in AI demand is placing unprecedented strain on existing data centre infrastructure, in terms of power, cooling, and space. At the same time, the shortage of skilled labour is slowing the pace of builds, upgrades and retrofits, accelerating the point at which facilities become inefficient or obsolete.

Almost a quarter (23%) of operators also cite supply chain disruption as a major concern. AI-ready data centres are particularly exposed to supply chain delays because they rely on globally sourced high-density compute, power, cooling systems and specialised cabling. Rising tariffs on components are squeezing tight budgets and forcing operators to re-evaluate sourcing strategies.

The report highlights several critical pressure points threatening the viability of data centre builds. AI demand has surged dramatically over the past year, with 92% of operators reporting a 42% average increase in AI workloads and 64% saying demand for AI-supporting capacity has exceeded expectations. This rapid growth is stretching both existing infrastructure and planned developments beyond their limits.

Compounding this pressure is the shrinking lifespan of current data centre models. Over half of operators (58%) believe AI will shorten the useful life of today’s facilities, as workloads quickly outpace available power, cooling, and bandwidth. As a result, 74% of operators are rethinking their strategies for power delivery, cooling systems, and site locations to ensure future AI-readiness.

One often overlooked but crucial vulnerability is cabling. The report finds that 70% of operators believe poor-quality cabling threatens long-term AI-readiness, while 27% highlight a shortage of skilled labour to install and maintain these systems. Without the right cabling infrastructure, data centres risk performance issues and diminished resilience over time.

A further 61% of respondents say they have experienced costly design reworks or retrofits due to siloed decision making during data centre builds. As a result, 70% of operators believe that holistic design principles will be critical for building AI-ready data centres. 

Commenting on the findings, Onnec’s Niklas Lindqvist said: “The AI race is global, but protectionist policies are forcing operators to make cost-based decisions that will limit the quality of infrastructure they can deliver. 

“With AI evolving so fast, operators need to make sure data centres are built to last – ensuring future sites have dynamic capacity when they’re up and running. With supply chains under pressure, sourcing essential components has become a major constraint.

Recommended reading

• Users Accidentally Leak Personal Data via Meta AI
• ChatGPT Search Inches Toward Strict EU Oversight as Users Surge
• AI Fuels Consumer Demand for Stronger Privacy Laws

“Coupled with a shortage of specialist engineers, operators may be forced to scale back or delay their ambitions, turning what were intended as cutting-edge facilities into compromised, less efficient builds. Meeting these challenges and easing cost pressures will be key to ensuring AI infrastructure is built to support long-term needs.

“Success in AI infrastructure will depend on far more than capacity alone,” continues Lindqvist. “AI is now the defining force in data centre strategy, but meeting that demand takes more than just scale – it takes foresight.

“Operators need to embrace holistic design thinking that considers every layer of the stack – from cabling and cooling to skills and sustainability. Failure to take this approach means operators risk building data centres that are already behind the curve the moment they go live.”