The 2026 US midterm elections are shaping up to be a watershed moment for artificial intelligence policy, as AI companies and their deep-pocketed backers deploy hundreds of millions of dollars to shape the political landscape. According to recent reports, the sector's spending through super PACs has reached staggering levels, with a remarkably consistent demand at the center of their efforts: a single, national framework for regulating AI, rather than a fragmented patchwork of state laws.
The industry argues that 50 different state regimes would stifle innovation, slow development, and hand a competitive edge to global rivals, particularly China. This push comes after a period of intense legislative activity at the state level, where over a thousand AI-related bills were introduced in 2025 alone, creating a complex and often contradictory regulatory environment that companies say is untenable for long-term planning and investment.
The primary vehicle for this political offensive is Leading the Future, a super PAC network launched in early 2025. Its initial funding includes massive contributions from prominent figures in the tech world: Marc Andreessen and Ben Horowitz, co-founders of Andreessen Horowitz, each reportedly donated $25 million, alongside OpenAI president Greg Brockman. The group's mission is explicit: to back candidates who support a "responsible national framework" for AI and to oppose those who undermine it. Already, the PAC has spent tens of millions of dollars on races spanning from Texas and Georgia to Illinois and Montana, targeting both primary and general elections.
But Leading the Future is not the only player. A newer organization, Innovation Council Action, has pledged approximately $100 million to support pro-deregulation candidates. Its reported donors span the intersection of crypto and AI wealth, including the Winklevoss twins and individuals connected to Elon Musk's xAI orbit. This influx of money underscores the high stakes for the AI industry, which sees the midterms as a critical juncture to lock in favorable regulatory conditions for years to come.
Why the money is flowing now
The timing of this massive spending is no coincidence. Washington's attempt to freeze state AI laws through a federal preemption provision failed spectacularly in the Senate, where the provision was stripped by a vote of 99 to 1 before the bill was signed into law. This defeat in Congress has shifted the battle to the ballot box. The industry is now betting that by electing friendly lawmakers, it can build a coalition that will pass federal legislation that overrides state-level rules before the 2026 election cycle concludes.
This strategy mirrors ongoing White House efforts to trade state preemption for federal online-safety rules, a deal that has yet to materialize. The gap between a stalled federal approach and a rapidly activating state house season is exactly the vacuum that this money aims to fill. The industry's message is clear: without a national standard, AI companies face compliance costs that could reach billions of dollars annually, potentially hobbling smaller startups and consolidating power among the largest incumbents.
The stakes are particularly high in states that have emerged as AI regulatory pioneers. For example, California, Connecticut, and Colorado have all passed or considered aggressive AI transparency and safety bills. Connecticut's proposed AI Bill of Rights, though not yet law, would impose strict obligations on developers and deployers. Texas has focused on AI in healthcare and criminal justice, while Illinois has led on biometric privacy. Each state's approach differs, creating a compliance nightmare for companies that operate nationally or globally.
Not every AI firm plays it the same way
The AI industry, however, is not a monolith when it comes to regulation. Anthropic, a leading AI safety company, reportedly contributed $20 million to political efforts, but with a caveat: the money was restricted to educating the public on AI policy rather than direct political campaigning. Anthropic has also publicly argued that governments should have the authority to block dangerous AI systems, a stance that sets it apart from many of its peers. This split highlights a broader tension within the industry between those who favor minimal regulation to spur growth and those who acknowledge the need for guardrails to prevent catastrophic risks.
Microsoft's president, Brad Smith, has separately complained that the US currently regulates AI "with no clear rules at all," a vacuum this spending would help fill on industry terms. For Microsoft, which has invested heavily in AI through its partnership with OpenAI, a national framework would provide the certainty needed to invest billions in data centers and research. But the devil is in the details: what counts as "light-touch" regulation for some may be too permissive for others, and the industry's political spending is designed to ensure that the final framework aligns with the interests of the biggest spenders.
The diversity of approaches extends to the types of candidates being supported. While Leading the Future primarily backs Republicans who favor deregulation, some contributions have gone to Democrats who have shown a willingness to engage with the industry. This bipartisan strategy reflects the understanding that AI regulation is not a purely partisan issue; senators from both parties have co-sponsored bills on AI transparency, election integrity, and copyright protection. The industry's goal is to find common ground on a framework that is both pro-innovation and sufficiently robust to appease public concerns.
The democracy question
Critics of this wave of spending see something more troubling than ordinary lobbying. Watchdog groups, including the Campaign Legal Center and Public Citizen, have branded the contributions as "dark money" and warned that corporations are trying to buy favorable regulation before the technology's full societal implications are understood. They argue that super PACs, even when they disclose donors, allow for influence that is disproportionate to the public interest, especially when the industry being regulated is as powerful and opaque as AI.
Supporters counter that political spending is legal, partially disclosed, and no different from other industries defending their interests. They point to the long history of pharmaceutical, energy, and financial sector PACs as precedent. The clash is sharpened by a public already wary of AI, evident in grassroots revolts that have blocked data-center projects in several states and cities. Citizens in places like New York, Virginia, and Arizona have protested against the energy and water demands of large AI facilities, raising questions about local control and environmental impacts that state-level regulation addresses.
More broadly, the debate touches on the fundamental question of who governs technology. In the absence of federal action, states have become laboratories of democracy, experimenting with approaches to AI that range from outright bans on deepfakes in elections to requirements for algorithmic impact assessments. The industry's push for federal preemption would effectively end these experiments, centralizing power in Washington and potentially weakening protections that states have designed to protect their residents. Whether that outcome is desirable depends on one's view of federalism and the risk of regulatory capture.
What the money buys, in the end, is not just advertising or campaign staff, but a frame for the debate: that AI rules should be national, light, and set with the industry in the room. This framing is already influencing the language of candidates and committees, who increasingly speak of "responsible innovation" and "American leadership" as code for the industry's preferred approach. Whether voters accept that frame is the test these millions are designed to pass.
As the 2026 midterms approach, the volume of AI-related political spending is likely to grow, with new super PACs emerging and existing ones expanding their reach. The outcome of these races will have long-term consequences not just for the AI industry, but for the safety, fairness, and transparency of systems that are increasingly woven into every aspect of daily life. The billions invested in lobbying and campaign contributions may prove to be one of the most consequential investments the tech sector has ever made.