The political fight over artificial intelligence infrastructure moved from statehouse hearings to television advertising this week. On 31 August 2026, Build American AI — an advocacy group affiliated with the AI-industry super PAC network Leading the Future — announced a multimillion-dollar campaign in election battleground states aimed at countering a rising public backlash against data centre construction.
The initial targets are Kansas, Ohio and Wisconsin, three states where local opposition to data centre projects has escalated into a statewide electoral issue. The group also said it is launching a new political arm to back candidates aligned with its agenda.
What the Campaign Involves
Build American AI is organised as a 501(c)(4) nonprofit and, according to reporting on the launch, has roughly $50 million cash on hand. The money is being directed at paid advertising, grassroots organising, research, earned media and public education, with a message that the United States should build the infrastructure needed to lead in AI while protecting families, communities and the environment.
Alongside the ad buy, the group is standing up a new 527 super PAC called Building the Future, which will support state and federal candidates favouring its policy positions and oppose those who do not. For the moment the organisation says it is focusing on issues rather than specific races. Building the Future has said it will back candidates supporting policies that preserve water resources and prevent electricity rate increases — an explicit attempt to absorb, rather than dismiss, the two grievances driving local opposition.
The campaign is expected to expand as state legislatures draft bills for their 2027 sessions.
The Money and the Distance
The funding network behind the effort is well documented. Leading the Future launched in 2025 with more than $100 million from stakeholders including Andreessen Horowitz, OpenAI President Greg Brockman and Palantir co-founder Joe Lonsdale, and reported raising $125 million in 2025. Andreessen Horowitz has contributed more than $115 million to federal candidates in the current midterm cycle, making it the largest donor tracked by The New York Times.
OpenAI as a company has publicly distanced itself from the network's operations. In a statement cited by TechCrunch, the company said it "does not direct the activities of LTF, or have visibility into their operations," and added that groups advocating on AI "should be clear about their policy views, be honest about whom they represent, and not use tactics like astroturfing."
That distinction — between a company's stated position and the political spending of its executives and investors — is likely to be a recurring theme as the campaign scales.
Why It Matters
Data centres have become the point where abstract AI policy meets concrete local politics. Zoning boards, utility commissions and county councils now hold effective veto power over the physical build-out that frontier AI depends on, and they answer to voters who experience the projects as noise, water withdrawal, land use and — most potently — electricity bills.
The scale of the friction is significant. Coverage of the campaign's launch cited polling showing a substantial majority of US adults opposed to data centres being built near them, and reported that at least 75 projects worth roughly $130 billion were delayed or blocked in the early part of 2026. Set against US technology firms planning to invest more than $700 billion in AI infrastructure this year, a sustained permitting slowdown is a material business risk, not a public-relations inconvenience.
This also marks a strategic shift in how the AI industry lobbies. Previous efforts by the same network focused on federal preemption — pushing Congress and the executive branch to override state AI rules, an agenda advanced through the December 2025 executive order establishing a national AI policy framework and the Justice Department's AI Litigation Task Force. Preemption fights are fought in courts and committee rooms. Data centre siting is fought in county elections. Moving into battleground-state advertising means the industry has concluded it cannot win the infrastructure question purely at the federal level.
The states cited as favourable models — Colorado, Georgia and Texas — illustrate the playbook: governors who have publicly championed data centre investment, and tax structures that exempt these facilities.
The Counter-Argument
Opponents will point out that a campaign funded by the primary beneficiaries of data centre construction is not a neutral information effort, and that promises about water and electricity rates carry more weight in enforceable permit conditions than in advertising copy. The astroturfing critique OpenAI itself raised in June cuts directly at 501(c)(4)-funded grassroots organising, where donor identities are not disclosed.
There is also a durability question. Advocacy spending can shift a single election cycle; utility rate structures, interconnection queues and water rights are governed by processes that outlast campaigns. Whether ad dollars translate into approved projects is genuinely untested.
What to Watch Next
Three markers matter over the coming months. First, whether Building the Future moves from issue advertising to endorsing named candidates ahead of the midterms. Second, whether opposing coalitions — utility ratepayer groups, environmental organisations and local government associations — assemble comparable funding. Third, and most consequentially, what the 2027 state legislative sessions produce: binding conditions on water use and ratepayer protection would suggest the backlash is shaping policy regardless of who spends more on advertising.
The AI infrastructure debate has become an electoral one. That is new, and it will not reverse.
