startups: AI has triggered the biggest gas-plant building boom in

The insatiable energy demands of artificial intelligence are driving an unprecedented surge in natural gas power plant construction across the United States, prompting clean energy advocates to pivot their strategy towards regulatory intervention. This boom, reported by the Associated Press, includes extending the operational life of aging coal plants, challenging existing commitments to emissions-free electricity and escalating power costs for consumers. As the rapid expansion of AI data centers outpaces the development of renewable alternatives, a critical battle is unfolding over grid access and energy policy that could shape the nation's power landscape for decades.

The AI Energy Paradox

The sheer scale of AI's energy consumption is staggering, with some data centers requiring more electricity than a medium-sized city. This immense demand presents a significant challenge for utility companies, which are finding it impossible to construct new wind and solar facilities at the pace required to meet these immediate needs. Consequently, natural gas has emerged as the default solution, triggering the largest construction wave in history for gas-fired power plants, and even leading to the deferral of scheduled closures for older, more polluting coal plants.

States Push Back with Clean Energy Mandates

In response to this growing energy crisis, several states are attempting to impose legal frameworks to safeguard their clean energy objectives. New York Governor Kathy Hochul is considering a bill that would mandate large data centers achieve at least 90% renewable energy by 2040, starting with benchmarks in 2030. State Senator Kristen Gonzalez, the bill's author, argues that these wealthy AI companies, capable of investing billions in data centers, should also finance their own clean power infrastructure.

Michigan, Oregon, and Minnesota have already passed legislation in the past 18 months, tying tax exemptions for hyperscale data centers to clean energy targets, requiring 90% clean energy within six years. Similar legislative efforts are reportedly underway in California, Illinois, New Jersey, Pennsylvania, and Virginia. This collective action highlights a growing concern that AI's energy appetite could derail established clean energy timelines.

The Regulatory Battle for Grid Access

Recognizing the difficulty of physically out-building the data center boom with new renewable projects, clean energy advocates are shifting their focus to influencing energy policy through regulatory channels. Their innovative tactic involves lobbying regulators to permit major power consumers to construct their own dedicated clean energy generation facilities and connect them directly to the existing grid. This approach seeks to decentralize renewable energy development and integrate it more directly with demand.

Colorado, for instance, has instructed Xcel Energy to establish such a program. In an April filing, Xcel acknowledged the potential customer benefits, citing successful Google initiatives connecting 115 megawatts of geothermal power in Nevada and 1,900 megawatts of wind, solar, and storage in Minnesota. Google's agreement with NV Energy is considered a pioneering model, with similar arrangements reportedly approved or pending in eight other states. The Corporate Energy Buyers Association (CEBA) has also secured a comparable deal with Georgia Power and is actively pursuing one in North Carolina.

The pitch to utilities is rooted in commercial pragmatism: they gain a substantial long-term customer who funds grid expansion, rather than risking losing that customer to entirely off-grid solutions. This strategy aims to leverage corporate investment in renewables directly into grid enhancement, rather than relying solely on utility-driven projects.

Why the Grid is the Decisive Front

The true battleground for the future of energy policy lies not in legislative chambers but in grid access and interconnection rules. Regulators have been expediting grid connections for data centers, effectively controlling the pace and type of new power infrastructure. The current bottleneck for AI development is no longer silicon but energy, with Nvidia-backed startups already raising significant capital to address data center power solutions.

Communities are also actively resisting the unchecked expansion of data centers, having successfully halted 75 projects valued at an estimated $130 billion in a single quarter. Federal attention is mounting too, with the House of Representatives considering a bill designed to shift the substantial energy costs associated with AI data centers back onto the technology companies generating them. CEBA's policy chief believes the decisions made regarding grid policy now are expected to dictate energy trends for the next two to three decades, even as new gas plants are rapidly being cemented into place.

FAQ

Q: Why are natural gas plants being built instead of more renewable energy sources to power AI data centers? A: The enormous and immediate electricity demands of AI data centers outstrip the current speed at which large-scale wind and solar projects can be developed and brought online. Natural gas plants offer a quicker, more readily available solution to meet this surging demand, leading to a record construction boom for these facilities.

Q: What steps are states taking to address the energy demands of AI and protect clean energy goals? A: Several states are implementing or proposing legislation to mandate clean energy benchmarks for large data centers. Examples include New York (90% renewable by 2040), Michigan (90% clean energy within six years for tax exemptions), and similar bills are in consideration in California, Illinois, New Jersey, Pennsylvania, and Virginia.

Q: What is the "regulatory backdoor" strategy clean energy advocates are pursuing? A: This strategy involves lobbying energy regulators to permit large power users, such as tech companies, to build their own clean energy generation (like solar or geothermal) and directly connect it to the existing power grid. This aims to integrate renewable energy production with demand more efficiently and leverage the financial power of tech giants for grid expansion, while also securing them as long-term utility customers.