State legislatures push targeted AI bills on children, employers and safety risks despite Trump's executive order threatening legal challenges to state laws.
Six months after President Donald Trump issued a direct warning to states against regulating artificial intelligence, at least a dozen state legislatures are moving ahead anyway. The defiance is orderly, not rebellious: lawmakers have read what happened to broader bills and come back with something harder to kill.
Congress has produced no federal AI legislation, leaving a vacuum that states are now rushing to fill. Their targets are specific: how chatbots interact with children, how employers use AI in hiring decisions, and what responsibilities developers carry when their systems contribute to large-scale harm. The shift reflects a growing recognition that artificial intelligence is no longer a future policy problem, according to Newsday.
Earlier ambitions collapsed. Several governors vetoed wider-ranging AI accountability bills in 2024 and early 2025, including measures that would have held developers responsible for algorithmic bias in their systems. Legislators recalibrated. The new wave of laws is narrower by design, targeting the specific corners of daily life where people encounter AI without knowing it.
Trump's opposition
Trump has made AI a top national and economic security priority, framing domestic regulation as a direct threat to U.S. competitiveness with China. In an executive order, he directed the attorney general to form a task force with one explicit mandate: challenge state laws deemed more than "minimally burdensome" to the industry. The Commerce Department was separately tasked with compiling a list of such laws. Howard Lutnick, the commerce secretary, and David Sacks, the White House AI and crypto czar, flanked him at the signing alongside Senator Ted Cruz.
His argument rests on scale. American companies are spending trillions of dollars building and deploying AI systems, and a patchwork of 50 different state regulatory frameworks could slow that investment during a race he considers existential. Critics from both political parties and consumer rights advocates pushed back hard: without a federal law in place, blocking state regulation amounts to a free pass for an industry that already operates with almost no binding oversight.
The federal vacuum
That critique lands harder against the actual legislative record. More than a year into the second Trump administration, Congress has not passed a single substantive AI bill. Executive-branch guidance and litigation threats are the only tools Washington has deployed.
The stakes are not abstract. ChatGPT reached 800 million monthly active users and $20 billion in annual recurring revenue by early 2026, making it one of the fastest-adopted products in consumer technology history, according to Forbes. Meanwhile, Meta spent more than $14 billion to recruit Alexandr Wang and build proprietary foundation models after its open-source strategy failed to attract paying customers, according to CNBC. Accountability structures for systems operating at that scale simply do not exist in federal law.
States, by default, have become the regulators of first resort.
What states are actually doing
Bills advancing now are deliberately narrow in scope. Rather than attempting to define or govern AI as a technology, they govern it in context: the chatbot a teenager uses for homework, the hiring algorithm a recruiter runs on thousands of resumes, the automated system that influences a decision affecting public safety. That precision makes federal preemption harder to argue and makes a governor's veto easier to oppose politically.
Whether Trump's task force will have the legal standing to strike down these more targeted laws remains genuinely unclear. Federal preemption of state consumer and labor law requires statutory authority that Congress has not granted for AI. Several state attorneys general have signaled they will contest any federal challenge, and Newsday reporting confirms that resistance spans both parties.
Historical echoes are hard to ignore. States led on data privacy after federal efforts collapsed through the 2010s, eventually producing the California Consumer Privacy Act and a slow cascade of state equivalents. Companies ultimately lobbied for a federal standard just to get regulatory consistency. Artificial intelligence regulation appears to be following the same arc, on a compressed timeline.
If Congress does not act, the U.S. may arrive at a de facto AI regulatory framework built entirely from state law, fragmented across jurisdictions and expensive to navigate. That outcome might ultimately be the pressure that breaks the federal deadlock, or it might simply become the permanent condition.
FAQ
Q: Can Trump legally block state AI regulations?
A: Not automatically. Federal preemption requires statutory authority that Congress has not granted for AI specifically. Trump's executive order creates a litigation task force, but court challenges from state attorneys general are expected and outcomes are uncertain.
Q: Which states are advancing AI legislation in 2026?
A: Pennsylvania is among the states named in current reporting. Multiple legislatures are targeting AI in employer hiring, AI interactions with minors, and systems capable of large-scale harm, though specific state tallies remain fluid.
Q: What does Trump's AI executive order actually require?
A: It directs the attorney general to identify and challenge state AI laws deemed more than "minimally burdensome" to industry development, and tasks the Commerce Department with compiling a formal list of such laws.
Q: Why has Congress failed to pass a federal AI law?
A: Federal AI legislation has stalled as of mid-2026. The absence of a national framework is precisely what has given states political cover to advance their own rules, repeating a pattern seen with data privacy regulation a decade earlier.
