Rising energy costs have driven a fifth of UK businesses to shift their AI workloads overseas, posing concerns for a government banking on AI for economic progress. A report by CUDO Compute highlights the hurdles British firms face, echoing sentiments across North America and Europe. Data collection by Censuswide reveals global power costs are recalibrating AI infrastructure strategies. Constraints such as planning delays and power connectivity issues are hampering expansion, notably in Santa Clara, California, where datacenters await power connections.

High energy prices in the UK, among the highest globally, are limiting AI scalability, threatening ambitions of making Britain a tech leader. Yet, firms balance the allure of regional compute sovereignty against rising costs. To tackle energy price hikes, the UK government plans measures including long-term contracts and increased levies to foster reliance on renewable energy sources over gas.

The US leads as a favorable AI hub, with India and Eastern Europe also viewed as promising. Despite infrastructure challenges, companies eye moving due to reliable energy access elsewhere, especially where government policies are inadequate to support cluster development. With energy now the primary cost driver in AI deployments, firms choose locations based on power availability, altering previously user-centric site selections. Cudo Compute reports firms are already scaling back or considering relocation due to energy concerns, emphasizing the necessity for rigorous AI deployment planning.

Companies are urged to integrate power acquisitions, zoning, and cooling systems into strategic core competencies rather than handling them as logistical afterthoughts.