The Green Party's data centre proposal marks a direct challenge to the UK's artificial intelligence ambitions and energy infrastructure during a period of acute resource scarcity. Zack Polanski, the party's technology spokesperson, has called for a temporary moratorium on new data centre construction, targeting the spiraling water and electricity consumption driven by AI training and inference workloads.
Polanski's intervention surfaces a genuine policy tension. Data centres powering large language models and generative AI systems consume vast quantities of both resources. A single training run for a major AI model can demand millions of gallons of water for cooling systems while drawing power equivalent to small cities. As tech companies including Google, Microsoft, and Amazon accelerate AI infrastructure buildout across Europe and the UK specifically, energy regulators and environmental groups have begun flagging sustainability risks.
The Green Party argues that existing regulatory frameworks fail to adequately constrain data centre expansion. Current planning rules do not systematically account for cumulative water depletion or grid strain from multiple facilities clustering in specific regions. Polanski's position calls for temporary restrictions that would force stricter environmental impact assessments before new projects receive approval.
This proposal arrives amid rising political pressure on AI governance in the UK. The government has positioned itself as "pro-AI" and growth-focused, resisting heavy-handed regulation. Yet water companies across England and Wales have declared environmental emergencies, with drought conditions persisting through 2024. The tension between growth rhetoric and resource reality has created space for Green Party intervention.
The data centre industry counters that infrastructure investment remains essential to UK competitiveness. Tech leaders argue that restricting new facilities simply pushes investment offshore to countries with fewer environmental protections. They also note that newer data centre designs employ more efficient cooling systems and renewable energy procurement than older installations.
However, Polanski's framing reflects broader leftist and Green critiques of AI expansion on environmental grounds. Campaigns have emerged across Europe demanding similar curbs. Germany's environmental movement has similarly targeted new data centre proposals, citing water stress. The concern transcends party politics in resource-constrained regions facing heat waves and supply chain pressure.
The proposal has limited immediate legislative prospects. The current Labour government favors AI growth and maintains close ties with tech sector leadership. Yet the debate signals that AI infrastructure cannot remain decoupled from climate and resource conversations in public discourse. Water scarcity compounds with energy transition challenges. Data centres powered by coal or aging gas infrastructure carry different environmental math than those on renewable grids.
Polanski's call ultimately reflects a growing recognition among some policymakers that unfettered AI expansion carries material costs beyond the balance sheet. Whether Westminster adopts temporary restrictions or pursues voluntary sector commitments remains an open question. What is clear is that data centre siting, environmental impact, and AI governance have entered mainstream political conversation in ways they had not two years prior.