OpenAI ends government non-disclosure agreements following overwhelming US midterm data center rejection
The Verge|AIDEN Editorial|Nov 06, 2026
Voters in the 2026 US midterm elections approved restrictions on data centers in 40 of 41 local ballot measures across 10 states. In response, OpenAI announced it will halt non-disclosure agreements with public officials and fully cover local power infrastructure costs.
The rapid expansion of artificial intelligence infrastructure has triggered a severe political backlash across the United States. According to reporting from The Verge, local voters considered 41 ballot measures regarding data center developments across 10 states during the 2026 midterm elections, approving bans or stricter permitting rules in 40 of them. North Dakota was the only location where voters chose not to impose new curbs, resulting in a 97.5 percent passage rate for regulatory restrictions. The wave of opposition stemmed from community concerns over electrical grid strain, rising utility bills, massive cooling water demands, and severe noise pollution. An NBC exit poll found that 56 percent of voters cited artificial intelligence and data centers as a major or key factor in their congressional voting decisions, prompting tech companies to quickly overhaul their engagement strategies.
Faced with mounting political pressure, OpenAI announced its National Community Compact and a policy framework for facilities it is rebranding as AI compute campuses on November 5, 2026. The central pledges include abandoning the long-standing industry practice of requiring non-disclosure agreements from local government officials during site negotiations and committing to cover all local power and grid infrastructure costs directly. OpenAI also pledged to release recurring reports detailing noise levels, water usage, and power consumption, while promising to curtail facility electricity use during peak hours. Chris Lehane, global head of public policy at OpenAI, framed the package as an effort to set a high baseline for artificial intelligence infrastructure development, seeking to address complaints regarding opaque negotiations and localized environmental burdens.
Despite the commitments, industry analysts and regulatory observers question the practical reach of the concessions. The policy framework applies exclusively to facilities directly owned and operated by OpenAI, leaving out the third-party cloud infrastructure where the company leases the vast majority of its operational compute capacity. Skeptics suggest the move functions largely as a defensive measure against impending congressional hearings and environmental legislation following shifting political ground in data center hubs like Texas. Over the coming months, developers and utility providers face an overhauled approvals landscape marked by mandatory public scrutiny, while major tech firms face mounting pressure to establish binding, industry-wide standards across both leased and proprietary facilities.
— AIDEN Editorial Team
What this means for the market
The voter-driven rejection of data center projects highlights a critical global shift from software safety debates to the physical constraints of energy and water distribution. As municipal resistance spreads beyond the United States, artificial intelligence developers and cloud infrastructure operators worldwide must prepare for prolonged project timelines, stricter public disclosure mandates, and higher capital outlays to cover power generation. Relying on opaque negotiations or leased third-party capacity will face intense scrutiny, accelerating the push toward legally mandated infrastructure standards across major technology hubs.
How this issue is unfolding
The expansion of massive data centers required to train and run generative artificial intelligence models sparked bipartisan public backlash across the United States due to grid congestion, rising consumer electricity costs, immense cooling water consumption, and noise pollution. During the 2026 US midterm elections, voters directly halted big tech infrastructure growth by approving data center bans or moratoria in 40 out of 41 local ballot measures across 10 states. The election results disrupted political dynamics in major computing hubs such as Texas and heightened pressure for congressional intervention, creating immediate obstacles for corporate expansion plans. In response, OpenAI introduced self-regulatory measures that rebrand facilities as compute campuses, eliminate non-disclosure agreements for public officials, commit the company to paying full regional grid costs, and establish mandatory environmental reporting. However, skepticism persists because the compact applies only to facilities directly owned and operated by OpenAI rather than leased cloud sites, fueling debate over whether the move is a temporary bid to evade regulation and whether binding standards should be legally enforced across the entire technology sector.
