US state legislators are defying the tech lobby to pass AI laws on safety and privacy, signaling a shift in power from Silicon Valley to local governments.

For years, the dominant narrative in the corridors of power in Washington and Silicon Valley was that artificial intelligence required a unified, federal approach. The argument was simple: a patchwork of fifty different state laws would stifle innovation and hand a competitive advantage to global rivals, specifically China. However, that strategy of centralized control is currently collapsing. State legislators across the United States are no longer intimidated by the threat of federal preemption or the deep pockets of the tech lobby.

The shift represents a fundamental change in political calculus. Where lawmakers once feared the electoral might of venture capitalists and AI giants, they now see the regulation of frontier models as a political asset. Public anxiety regarding the rapid deployment of AI—ranging from the erosion of privacy to the displacement of workers—has reached a tipping point, turning AI oversight into a winning campaign issue during an election year.

The failure of the federal preemption strategy

The attempt to freeze state-level action reached a critical juncture in mid-2025. Allies of the Trump administration, led by figures such as Senator Ted Cruz, attempted to insert a provision into a major Republican bill that would have barred states from regulating AI for a decade. The goal was to create a regulatory vacuum at the state level, ensuring that only federal guidelines governed the industry.

This move backfired. Instead of intimidating statehouses, the threat of a ten-year freeze acted as a catalyst, forging unusual bipartisan coalitions between red and blue states. The Senate eventually voted down the provision, marking a significant defeat for the tech industry's strategy of federal exclusivity. As Politico reports, this failure has emboldened local legislators to craft measures specifically designed to be immune to future federal attempts at preemption.

Money vs. Mandates: The decline of the tech lobby

The financial scale of the opposition to state laws was immense. A political network known as Leading the Future, fueled by over 0 million from OpenAI President Greg Brockman and venture capitalists Marc Andreessen and Ben Horowitz, sought to support candidates who opposed state-level AI restrictions. The strategy was to use financial leverage to ensure that state legislators remained passive.

However, the influence of this capital is waning. In Utah, Representative Doug Fiefia provides a stark example of this reversal. Fiefia had previously abandoned a bill focused on AI and children's safety following pressure from the White House and industry lobbyists. After defeating a state Senate incumbent who was backed by the Leading the Future network, Fiefia announced his intention to revive the proposal. The message is clear: the fear of Silicon Valley's spending is being replaced by the fear of voter backlash.

Priority targets for new state legislation

Statehouses are not merely passing symbolic gestures; they are targeting the structural core of how AI is developed and deployed. The current wave of legislation focuses on several high-friction areas that the tech industry has historically avoided regulating.

Lawmakers are prioritizing the safety of frontier models and the implementation of independent audits to ensure these systems do not pose systemic risks. There is also a concentrated effort to protect minors, with Rhode Island already enacting chatbot safeguards in June 2026. Beyond the software itself, the physical infrastructure of AI is coming under scrutiny. The environmental impact and economic strain caused by massive AI data centers have become flashpoints for local opposition, leading to bills that address the resource consumption of the hardware powering these models.

The landscape has changed dramatically, and lawmakers are increasingly viewing the fight against the industry as an electoral advantage.

Massachusetts is currently exploring an AI safety bill that could potentially be more stringent than the existing frameworks in California, New York, and Illinois. This suggests a race to the top in terms of regulatory rigor, rather than a race to the bottom to attract tech investment.

A fragmented landscape for AI developers

The result of this defiance is the very thing the tech lobby warned against: a fragmented regulatory environment. For entrepreneurs and AI companies, this means the cost of compliance is rising. Instead of adhering to one set of federal rules, firms must now navigate a complex map of state-specific mandates. Politico Pro indicates that lobbyists who once tried to freeze out the states are now rushing to collaborate with them, realizing that participating in the drafting process is better than facing a finished law that ignores industry realities.

This shift mirrors other areas of tech regulation where states have led the way before federal action, such as data privacy in California. However, the speed and scope of the AI movement are unprecedented, leaving many companies struggling to keep pace with the legislative cycles of multiple state governments.

The broader trend of institutional pushback

The defiance seen in AI regulation is part of a larger global trend where institutional bodies are reclaiming ground from disruptive tech. This is evident not only in AI but also in the financial sector. For instance, the G20 has recently moved to address the frictions in cross-border payments—the very gaps that stablecoins exploited to gain a B2B edge. By pushing for expanded operating hours for payment systems and the adoption of the ISO 20022 data model, the G20 is attempting to modernize the traditional banking system to neutralize the competitive advantage of digital assets.

Whether it is the G20 targeting the sales pitch of stablecoins or US states targeting the safety of LLMs, the pattern is the same: the era of 'move fast and break things' is being replaced by an era of institutional orchestration. The PYMNTS analysis highlights that as polling becomes more negative toward unregulated AI, the political incentive to act outweighs the economic incentive to remain permissive.

Global implications for international enterprises

For businesses operating in the USA, UK, and global markets, the collapse of the federal preemption strategy in the US creates a high-risk environment for AI deployment. Companies can no longer wait for a 'single US law' to guide their compliance strategy. Instead, they must adopt a 'highest common denominator' approach, where they build their systems to meet the strictest state standards—likely those of California or Massachusetts—to ensure viability across the entire US market.

In the UK and EU, where the AI Act and other frameworks provide a more centralized (though complex) structure, the US situation introduces a new variable. US-based AI providers may find their products restricted in certain states, affecting the global availability of their tools. For international entrepreneurs, this means that 'US compliance' is no longer a binary checkmark but a continuous monitoring process of state-level legislative sessions. The shift from Washington-centric power to state-centric power means that local political sentiment now dictates the boundaries of global tech innovation.

FAQ

---

Sources: Pymnts, Subscriber, Politico ·