As artificial intelligence (AI) data centers expand rapidly across the United States, concerns over their impact on energy prices and local infrastructure are mounting. While at least six states have introduced legislation to slow or restrict new data center construction, the White House under President Donald Trump is exploring a less rigid approach. Instead of federal mandates, the administration is considering a voluntary agreement with major AI companies to address the challenges posed by these energy-intensive facilities.
In response to rising electricity costs and increasing consumer pushback, several states, including New York, Maryland, Georgia, Oklahoma, Virginia, and Vermont, have introduced measures to limit the proliferation of AI data centers. For example, New York lawmakers have proposed a three-year moratorium on new data center construction to develop policies that could mitigate their effects on household utility rates. However, the Trump administration has taken a different path by proposing a voluntary compact that would seek industry cooperation without imposing binding regulations.
According to a draft proposal obtained by Politico, the White House aims to have leading technology companies and data center developers commit to a set of principles addressing energy use, water consumption, and local community impacts. A White House spokesperson noted that the effort is aligned with ensuring Americans do not "pick up the tab" for the growing demand for power by data centers, signaling that further announcements may follow.
The draft compact outlines several objectives designed to minimize the burden on energy grids and local consumers. Among the key provisions, data center developers would be expected to:
The overarching goal, as stated in the draft, is to ensure that local ratepayers "are held harmless" and, ideally, experience reduced electricity prices in areas where large AI data centers are located.
Beyond energy pricing, the compact also highlights commitments to address water usage in regions experiencing scarcity. It calls for companies to become "water positive" and to coordinate with federal, state, and local regulators for grid planning and reliability. Additionally, companies would agree to curtail data center operations during peak demand periods and use backup power generation during grid emergencies. They would also be tasked with mitigating other local impacts, such as noise, traffic, and environmental disruptions.
The White House’s reliance on voluntary industry cooperation stands in stark contrast to the more rigid measures being pursued by state legislators. As states explore moratoriums, permitting bans, and mandatory rate protections, the federal strategy focuses on preserving the rapid pace of AI infrastructure development while encouraging companies to internalize their costs through negotiated commitments.
Whether a voluntary compact can provide the same level of consumer protection as binding state laws remains an open question. The decentralized nature of the U.S. electricity grid and the pivotal role of state utility regulators in setting rates add complexity to achieving nationwide consistency.
The draft compact underscores an evolving perspective on AI infrastructure. Once seen primarily as an economic development opportunity, the rapid expansion of data centers is now being scrutinized through the lens of consumer protection and cost allocation. As the White House seeks to strike a balance between industry growth and public interest, the administration’s approach may ultimately depend on whether states and regulators find voluntary agreements sufficient to address their concerns. For now, the proposal represents one of the most ambitious efforts to shape the footprint of AI-driven infrastructure without imposing direct federal regulations.
