AI data centers in California are now entering a strict regulatory framework, as Gov. Gavin Newsom signed seven bills aimed at protecting residents from rising energy and water costs caused by the explosive growth of artificial intelligence. The new legislation imposes transparency, financial responsibility and environmental compliance obligations on all companies that operate or plan to operate data centers in the state.
See also: Intel and AMD "lock" multi-year agreements with Chinese data centers
The move comes at a critical juncture, as demand for computing power from artificial intelligence companies like OpenAI, Google, Meta and Microsoft has skyrocketed, putting enormous strain on California’s power grid and water resources. According to estimates, a large data center can consume as much water as a small city, while its energy consumption can reach hundreds of megawatts.
The package of laws includes regulations for both existing facilities and new ones being planned, and is one of the most comprehensive efforts of any U.S. state to regulate the impact of artificial intelligence on local infrastructure.
Data Centers AI: What the new California bills provide
The centerpiece of the legislative package is the California Public Utilities Commission (CPUC) to introduce a new pricing category exclusively for data centers. This means that data centers will no longer be able to benefit from the same tariffs as residential consumers or small businesses, but will pay a separate, higher tariff that reflects their actual consumption and impact on the network.
Additionally, data centers AI will be required to fund upgrades to local power grids and water systems that are necessitated by their operations. Until now, these costs have often been passed on to local consumers through increases in electricity and water bills — a practice that Newsom called unjustified and unfair to residents.
Another critical element of the legislation is the disclosure requirement. New data centersin the planning stages will be required to disclose to local authorities estimates of water consumption, energy efficiency and their plans for droughts. This transparency will allow local communities to assess the impact of new facilities in advance and make informed decisions.
See also: AWS data centers: Drone attacks caused extensive damage
AI Data Centers and Environmental Requirements: The New Era of Regulatory Compliance
To gain access to a simplified permitting process, AI data centers will have to meet specific requirements in three areas: energy, water, and fuel. This is essentially a “carrot” for companies seeking faster permitting — but only if they can demonstrate that they are operating responsibly. The logic is simple: the more efficient and environmentally friendly a data center, the faster it can receive the necessary permits.
The need for these regulations is understandable when you consider the scale of the stakes. According to industry estimates, global data center energy consumption is expected to double by 2030 , with AI data centers being the fastest growing sector. In California, the situation is particularly pressing due to the combined pressure of climate change — which is intensifying droughts — and the increasing demand for computing power.
It’s worth noting that data center cooling is one of the largest consumers of water in modern industry. Cooling systems that rely on water evaporation — known as evaporative cooling — can consume millions of liters of water per day. In a state that faces chronic drought problems, this isn’t just a financial issue — it’s a matter of survival for local communities.
Governor Newsom was clear in his statement: “As the Trump administration moves toward deregulation, communities are being called upon to deal with the consequences — higher electricity demand, grid constraints, water consumption and pollution. With these laws, we are ensuring that Californians remain in control — and that those who profit from data centers do not do so at our expense.”
This statement reflects a broader trend: while the federal government under Trump has taken a liberal stance on regulating artificial intelligence, many states — and California in particular — are moving in the opposite direction, seeking to enact strong regulatory frameworks that protect their citizens.
That same week, Newsom also signed an executive order to expedite the creation of a “kill switch” for AI — an emergency mechanism that would allow for the immediate shutdown of dangerous or unregulated AI systems. This shows that California is not treating AI data centers just as infrastructure, but also as a source of potential risks that need to be actively managed.
The industry’s reaction to the new bills has been mixed. Some tech companies have welcomed the clarity the new regulatory framework provides, while others have expressed concerns about increased costs and potential delays in licensing new facilities. Some industry analysts have noted that the regulations could lead companies to choose other states for their new facilities — although California remains one of the most attractive destinations due to its human resources and technological infrastructure.
The issue of AI data centers and their impact on local infrastructure is not unique to the US. In Europe, and particularly in countries like Ireland and the Netherlands, similar concerns have led to tighter regulations and even temporary bans on new data centers in some areas. Greece, which is emerging as one of the most attractive data center in Southeast Europe — with major investments from companies like Microsoft — will also need to develop a strong regulatory framework that balances economic growth with the protection of local resources.
See also: Tribeca Data Leak – Angelina Jolie, Robert De Niro and other Hollywood stars affected
Overall, California’s new AI data center are an important step toward a more responsible and sustainable development of artificial intelligence. They set clear rules for who pays for the digital revolution and ensure that local communities are not unfairly burdened by the profits of big tech companies. If this model proves effective, it could very well become a model for other states and countries facing similar challenges.
