The Backlash Goes Structural

On April 15, Maine's legislature passed the nation's first statewide moratorium on large-scale data centers and sent it to the governor's desk. The bill would freeze new development for 18 months — until November 2027 — while a coordinating council studies the infrastructure, environmental, and utility impacts of AI computing facilities. As of this writing, Governor Mills has not signed it. She requested an exemption for a proposed data center at a retired paper mill in Jay; the legislature rejected the carveout.

Maine is not an outlier. It's the leading edge. According to Data Center Watch, $64 billion in data center projects have been blocked or delayed nationwide over the past two years — $18 billion killed outright, $46 billion stalled. A total of 142 activist groups across 24 states are organizing against construction and expansion. The opposition is bipartisan: Republican officials cite tax giveaways and grid strain; Democrats cite environmental damage and resource consumption.

The pattern is consistent. In Festus, Missouri, voters replaced half the city council this month over a data center project. In Ohio, residents are gathering 400,000 signatures for a ballot measure to permanently ban hyperscale data centers. In Prince William County, Virginia, a $25 billion project is stalled by multiple lawsuits. In Warrenton, Virginia, every council member who voted to approve an Amazon data center has since lost re-election. In Indianapolis, a city councilmember who backed a data center had 13 bullets fired into his home — with a "No Data Centers" note on the doorstep. His eight-year-old son was inside.

The opposition is not anti-technology. It is anti-extraction — communities refusing to bear the costs of infrastructure that primarily benefits out-of-state corporations. Higher utility bills, strained water supplies, noise, land use decisions made without meaningful local input. Polls show 65 percent of Americans oppose new data center facilities near their homes.

150 Bills, One Preemption Clock

The Stanford AI Index, released April 13, confirmed what state legislatures already know: states passed a record 150 AI-related bills in 2025 while the federal government stalled. The report's assessment is blunt — regulation is running behind the technology because, in the words of the Index's director, "we don't really understand how it works."

The White House is moving in the opposite direction. The March 20 National Policy Framework recommends that Congress establish federal preemption of state AI laws, arguing that a "patchwork of 50 different regulatory regimes" makes compliance burdensome for AI companies. The December 2025 executive order established an AI Litigation Task Force empowered to challenge state laws in court, directed the Commerce Department to identify "onerous" state regulations, and conditioned BEAD broadband funding on states not enforcing disfavored AI laws.

Congress has repeatedly declined to enact broad preemption — stripping it from the One Big Beautiful Bill Act and the National Defense Authorization Act. Senator Blackburn's 291-page TRUMP AMERICA AI Act packages preemption alongside child safety and copyright provisions, but it has not been formally introduced. Meanwhile, Senator Sanders introduced the AI Data Center Moratorium Act in March — pushing the opposite direction entirely. The federal response is now fractured: preemption from the executive branch, moratorium proposals from the legislative branch, and no framework from either. Legal analysts at multiple firms note that the executive order alone cannot preempt state law — only Congress can do that — but the administration's strategy of threatening funding, launching litigation, and creating regulatory uncertainty may have a chilling effect regardless.

The Privacy Law That Isn't an AI Law

Governor Stitt signed SB 546 on March 20, making Oklahoma the 20th state to enact a comprehensive consumer data privacy law. Effective January 2027, it gives Oklahomans the right to access, correct, and delete their personal data, and to opt out of targeted advertising.

The law is modeled on Virginia's and Tennessee's business-friendly frameworks. It does not regulate AI decision-making. An Oklahoman can now opt out of a targeted ad. They cannot find out when an AI system denies their insurance claim, screens out their job application, or flags their medical records.

Colorado's draft AI Act overhaul — released the same week — requires consumer notification when automated systems influence major life decisions and provides a right to human review. Oklahoma's privacy law has no equivalent provision.

SB 1521 Advances — Part of a National Wave

Oklahoma's chatbot safety bill, SB 1521, cleared a House committee after passing the Senate 43-0. The bill prohibits AI chatbots designed to simulate emotional relationships with minors and requires age verification. Nebraska, Oregon, Idaho, and Maine have passed similar legislation this session.

This is meaningful, bipartisan, and responsive to documented harm. It is also, like every other Oklahoma AI bill this session, a response to a specific harm rather than a systemic framework.

The Counter-Argument: "Existing Law Is Enough"

The April issue of the Oklahoma Bar Journal published an article arguing that Oklahoma's existing corporate law framework — fiduciary duties, the business judgment rule, contract law — already provides sufficient guidance for AI governance. Corporate boards can rely in good faith on AI outputs the same way they rely on expert reports.

The argument is legally sound and practically incomplete. It addresses the obligations of corporate boards to their shareholders. It does not address the obligations of corporations to the citizens affected by their AI systems. An Oklahoma business that uses AI to screen job applicants can satisfy its fiduciary duties while systematically discriminating against protected classes — and under current law, the applicant has no AI-specific right to know.

The Bar Journal article is the strongest version of the "we don't need new law" position. But it answers the wrong question. The question isn't whether corporate boards can govern their own AI use. The question is whether citizens can govern the AI that governs them.

Colorado Rewrites Its AI Act

Governor Polis released a draft bill to replace Colorado's 2024 AI Act. The new version requires companies to notify consumers when automated systems influence major life decisions — loan approvals, employment outcomes, insurance determinations. If a decision goes against a consumer, they have the right to see the reasoning, correct errors, and request human review.

Colorado's approach is imperfect and already drawing industry opposition. This month, xAI filed a lawsuit challenging the law directly — the first major legal test of a state's comprehensive AI governance framework. But Colorado represents something Oklahoma has not attempted: a framework that treats AI governance as a citizen rights issue, not just a consumer protection or child safety issue.

Oklahoma has the bipartisan consensus to build something like this. It has the votes. What it doesn't have is the bill.