President Donald Trump is reportedly considering a draft executive order aimed at pressuring states to cease their regulation of artificial intelligence, a move that parallels discussions among some members of Congress to temporarily prohibit states from enacting AI laws. Trump and certain Republicans contend that existing state regulations, along with future ones, could stifle innovation and economic growth within the rapidly expanding AI sector. Conversely, critics from across the political spectrum, including civil liberties and consumer rights organizations, express concern that a ban on state-level regulation would unduly benefit large AI corporations by limiting oversight.

Currently, four states—Colorado, California, Utah, and Texas—have implemented laws that establish rules for AI usage across the private sector, according to the International Association of Privacy Professionals. These regulations typically focus on restricting the collection of specific personal information and demanding greater transparency from companies developing and deploying AI systems. The urgency for such laws stems from the pervasive nature of AI in daily life, where it increasingly influences critical decisions for Americans, such as eligibility for job interviews, apartment leases, home loans, and even specific medical treatments. Research has consistently highlighted AI's propensity for error in these decisions, sometimes exhibiting biases that prioritize particular genders or races.

Calli Schroeder, director of the AI & Human Rights Program at the public interest group EPIC, emphasizes the distinction between human and AI decision-making. While a human can be asked to explain their conclusions and the factors considered, eliciting such explanations from an AI system is often impossible, even for its programmers. States with more ambitious regulatory proposals aim to compel private companies to provide transparency and rigorously assess the potential for discrimination within their AI programs. Beyond these broad rules, many states have targeted specific applications of AI, such as prohibiting the use of deepfakes in elections and in the creation of nonconsensual pornography, as well as establishing guidelines for governmental AI deployment.

The leaked draft executive order outlines a plan for federal agencies to identify state AI regulations deemed 'burdensome' and subsequently pressure states to revoke or avoid enacting them. This pressure could manifest through various means, including the withholding of federal funding or legal challenges to state laws in court. Furthermore, the order would initiate a process to develop a national, 'lighter-touch' regulatory framework intended to supersede existing state AI laws. Trump's rationale behind this approach is that a fragmented landscape of 50 different state regulations impedes the growth of AI companies and potentially allows rivals like China to gain an advantage in the global AI race. He has also expressed concerns about state regulations leading to what he terms 'Woke AI.'

While the draft executive order is subject to change, with a senior Trump administration official indicating a tentative plan for its signing, House Republican leadership is also exploring a separate proposal to temporarily block states from regulating AI. House Majority Leader Steve Scalise has confirmed these discussions, though the specifics of such a proposal and which AI regulations it would override remain unclear. TechNet, a lobbying group representing major tech companies like Google and Amazon, has previously argued that a pause on state regulations would particularly benefit smaller AI companies in their nascent stages and provide lawmakers with crucial time to craft a nationwide regulatory framework that effectively balances innovation with accountability.

However, attempts at federal intervention to ban state AI regulations have previously met with failure, partly due to opposition from within the Republican party itself. Florida's Republican Governor, Ron DeSantis, publicly criticized a federal ban on state AI regulation as 'Not acceptable' on X, arguing that such a move would constitute a 'subsidy to Big Tech.' DeSantis further contended that it would prevent states from implementing protections against various harms, including 'predatory applications that target children' and 'online censorship of political speech.' Cody Venzke, senior policy counsel at the ACLU’s National Political Advocacy Department, also noted the unpopularity of a federal ban on state AI regulation, stating that the American public desires AI to be trustworthy, not discriminatory, unsafe, or prone to 'hallucinations.' This contentious debate highlights the complex challenge of fostering technological advancement while simultaneously safeguarding public interests and ensuring ethical AI deployment.