California shifts data center costs to operators as US projects face delays
Governor Gavin Newsom signed seven laws requiring data center operators to cover infrastructure costs and undergo full environmental reviews. This state-level tightening contrasts with federal efforts to remove construction barriers, even as 47 large US projects face delays or cancellations.
California Governor Gavin Newsom signed seven new laws regulating the data center industry, marking a significant shift in the state's approach to AI infrastructure. Three of the laws mandate that data center operators, rather than residential ratepayers, bear the costs of electric infrastructure upgrades. Three additional laws require the disclosure of water usage and other resource consumption, while the seventh law removes data centers from environmental review exemptions. Unless specific conservation standards are met, these facilities must now undergo standard environmental review processes. This regulatory stance stands in direct contrast to the federal government's position under President Trump, which aims to eliminate blockers to data center construction. Industry critics, including data center building companies, have argued that the new requirements will increase costs and extend timelines due to the duration of environmental reviews. The legislation reflects growing public concern over the environmental and economic impacts of AI infrastructure in California, despite lobbying efforts from major tech companies headquartered in the state. Meanwhile, the broader US market is experiencing headwinds. Forty-seven large construction projects in the United States were delayed, canceled, or blocked this year. These setbacks are attributed to community opposition, stricter local regulations, and the fact that current revenue from AI applications does not yet cover high construction costs. Rising interest rates and inflation have further unsettled investors in the sector. In a separate development, the German federal government announced plans for close cooperation with NVIDIA focused on physical AI in industrial manufacturing. This initiative aims to connect AI models with real machines, such as automated plants and humanoid robots, and will be supported by a three-digit million euro budget, with NVIDIA providing the necessary compute units.