Currently, only 20% of data centers in Europe and the Middle East are optimized for AI requirements, despite escalating infrastructure demands for AI processing acceleration. According to a data center consultancy BCS report, capacity suited for AI is limited, with significant challenges for developers, particularly a skills shortage.

The report anticipates AI-ready facilities might rise to around 70% by 2030, although infrastructure demands could still surpass available resources. These insights, derived from over 3,000 industry experts across 41 countries, highlight current constraints that delay new facility construction in Europe.

Primarily, existing sites were built for typical enterprise or cloud use and are not configured to support the higher power densities and stringent cooling needs of AI servers. While locations might seem capacitated on paper, they are often unable to manage sustained power and cooling requirements.

The 20% readiness figure is less an industry indictment and more reflective of how demanding AI infrastructure has become. Developers face combined challenges of scarce land, inadequate materials, power connectivity issues, and skilled workforce shortages.

BCS’s chief, James Hart, noted that these are not isolated challenges but frequently co-occurring barriers. As market growth expectations remain high, the ability to meet these projections is becoming increasingly restricted.

Market data indicates a strong demand for increased data center capacity, yet supply is tightening. Professionals foresee a growing skills shortage and assert that supply chain volatility is becoming a perpetual challenge. Many developers have missed deadlines due to these compounding issues.

Despite forewarnings, the UK anticipates an increase of more than 10 GW over the next decade, with significant expansions also expected in Germany, Austria, and Switzerland. Even so, London’s data center supremacy may be succeeded by Frankfurt by 2031.