California and EU introduce new rules on AI data center resource use

Solmar Insights

California and the European Union have unveiled new regulatory measures targeting the environmental footprint of AI data centers this week. The rules focus on monitoring and reducing water and energy consumption as digital infrastructure demand accelerates.

Regulation scope and intent

The latest rules announced by California authorities and the European Union are part of a growing policy response to the significant resource demands of hyperscale and AI computing infrastructure. While detailed legislative text was not released alongside the announcement, both regulatory actions signal intent to collect operational data on data center water and electricity usage and, over time, set enforceable limits targeting resource-intensive AI workloads.

This regulatory push comes as operators report increased resource draw from AI training clusters and large language models, resulting in higher grid peak demands and greater cooling requirements. The emphasis of these rules on both energy and water use reflects the recognition that next-generation AI data centers are driving up not only electrical loads but also onsite and municipal water consumption, especially for evaporative and liquid cooling systems.

Market drivers and infrastructure implications

Regulators’ focus stems from digital infrastructure’s outsized share of growth in grid connections and water sourcing. The surge in AI-driven demand for compute has created a new layer of scrutiny for data center deployments in California and EU member states, regions often stressed by water scarcity and grid capacity limits. Infrastructure operators face challenging requirements to deliver power-dense AI platforms without breaching local or state sustainability mandates.

Transparency obligations, now looming on both sides of the Atlantic, could drive further standardization of environmental reporting for hyperscalers, colocation providers, and sovereign cloud deployments. For grid operators and utility planners, these data center-centric rules could feed into future distribution and transmission planning, as well as inform demand response or site prioritization for new projects.

Risks and compliance challenges

The compliance burden for data center operators is rising, particularly for facilities running advanced AI hardware. Reporting metrics will likely include real-time energy use, source mix, and water withdrawal, driving new investment in metering and monitoring solutions. The dual regulatory activity across California and the EU increases the complexity for multinational operators, who must synchronize disclosures and compliance across different jurisdictions and reporting timelines.

Energy and water use tracking could also impact developers in the siting and permitting phase, with more value placed on locations offering sustainable grid capacity and non-potable water options. As both California and the EU account for a significant share of new AI data center projects, these regulations could affect procurement strategies, partner selection, and even future contract terms for land and interconnection.

Potential effects on project pipelines

Developers and investors may now face tighter timelines for resource disclosure and potentially tougher approval environments, especially as enforcement mechanisms mature. The regulatory clarity on both sides could accelerate the adoption of newer cooling methods, such as direct liquid cooling, and renewable or recycled water sourcing, particularly in urban and drought-prone regions.

For hyperscalers seeking to expand AI computing footprints in the United States and Europe, structured rules may also give rise to new sustainability-linked financing or lease terms. Meanwhile, competing U.S. regions that have not implemented similar regulation may see increased data center siting activity as operators weigh compliance costs versus speed to deployment.

Reporting via the original publisher