High Energy Costs Stall OpenAI’s UK Data Centre Plans

OpenAI has paused plans for a major UK data centre project, highlighting the growing role of energy in shaping the future of artificial intelligence infrastructure.

The proposed “Stargate” project—part of a wider push to build large-scale AI computing capacity—has been put on hold due to high electricity costs, grid constraints, and regulatory delays. The company is expected to revisit the investment when conditions become more favourable.

The pause also reflects a broader global strategy. Stargate is designed to deliver gigawatt-scale AI infrastructure across multiple regions, with investment increasingly directed toward locations where energy is abundant, reliable, and quick to access. In practice, this is pushing projects toward parts of the United States, the Middle East, and Asia, where faster deployment and stronger power availability offer a clear advantage.

At the heart of the issue is energy. Modern AI data centres are extremely power-intensive, running vast clusters of processors around the clock to train and operate models. These facilities require constant, high-volume electricity, often on the scale of small cities, making energy cost and reliability critical factors in where they are built.

This growing demand is beginning to shape national strategies. The UK government has set out plans to significantly expand domestic AI computing capacity and develop dedicated AI Growth Zones with improved access to power and streamlined planning.

The decision highlights a wider challenge for the UK. As demand for AI accelerates, energy availability, infrastructure, and speed of deployment are becoming just as important as technological capability in attracting investment.