What are Britain’s AI growth zones and are the plans feasible or ‘complete bunk’?
TL;DR
Britain has announced five AI growth zones: areas where companies are meant to build AI datacentre complexes of 500MW or more by 2030. The Guardian found major gaps in Lanarkshire. The £8.2bn CoreWeave and DataVita project was sold as powered by on-site renewables, but the government now says it will connect to the grid. DataVita points to more than 1GW of new energy, yet public planning records appear to cover only a fraction of the land needed. UK grid connections can take eight to ten years.
Nauti's Take
This looks like infrastructure policy written for the press release first. An AI growth zone is only meaningful if grid access, land, energy contracts and construction timelines are credible before the announcement.
Lanarkshire, as reported here, reads like the opposite: huge job numbers, huge renewable claims, thin evidence underneath. A serious AI plan starts at the substation, not at the launch event.
Briefingshow
AI infrastructure does not fail on strategy slides; it fails on power, land, permits and grid access. If governments announce 500MW clusters without a credible energy path, they create a political queue, not a compute advantage. The wider lesson for Europe is blunt: AI sovereignty needs infrastructure realism, not just GPU storytelling.
Sources