Trump signed an executive order Thursday that aims to create a “single national framework” for AI.
AUSTIN, Texas — President Donald Trump has signed an Executive Order that aims to restrict states from regulating artificial intelligence. Texas is one of a handful of states that have passed laws setting rules for AI across the private sector.
The order comes just days before a significant piece of legislation regulating AI in Texas is set to take effect. It would, among other things, limit how state governments could use AI and crack down on the use of AI for child pornography.
Trump signed the executive order in the Oval Office on Thursday. It directs federal agencies to identify burdensome state AI regulations and pressure states not to enact or retain them, either by withholding federal funding for programs like broadband or by challenging the state laws in court. It has a carve out for laws related to AI and child safety.
“My Administration must act with the Congress to ensure that there is a minimally burdensome national standard — not 50 discordant State ones. The resulting framework must forbid State laws that conflict with the policy outlined in this order,” the order said. “That framework should also ensure that children are protected, censorship is prevented, copyrights are respected and communities are safeguarded. A carefully crafted national framework can ensure that the United States wins the AI race, as we must.”
Trump said that, with the executive order, he hopes to create a “single national framework” for AI.
“We want to have one central source of approval,” Trump said. “Every time you make a change, and it could be a very reasonable change, you still won’t get it approved if you have to go to 50 states, so this centralizes it.”
Trump said the regulations hinder innovation and growth for the technology, but critics argue that banning state regulations would allow big AI companies to run wild with little oversight.
With no comprehensive federal legislation regulating AI, the U.S. has a patchwork of regulations that vary state by state, which Trump said gives China a significant advantage in the AI arms race.
“Globally, we’re having to contend with both the desire to be leading in the development of AI and also a deep understanding that we have to be responsible in the way that we proceed with AI, and so different countries are taking different approaches. Some of them are more restrictive than others,” Samantha Shorey, an assistant professor in the Department of Communication, and a fellow at the Roosevelt Institute at the University of Pittsburgh, said. “Figuring out how to balance the need for innovation alongside the need for responsible development, I think, is the puzzle that’s facing all governments right now.”
Shorey said much of the regulation currently at the state level is what she considers “common sense regulation of AI.”
“It’s difficult for me to understand what that does to stop the development of AI if citizens just need to be informed when a company or when a state government is using AI to communicate with them,” Shorey said. “Much of the successful legislation we’ve seen has been around transparency and just basic information about when AI is being used.”
Texas’ major AI regulation law is set to take effect on Jan. 1. The Texas Responsible Artificial Intelligence Governance Act, or TRAIGA, outlaws the use of AI for manipulating human behavior, like encouraging people to harm themselves or others, violating someone’s constitutional protections, unlawful discrimination against a protected class and generating child pornography or explicit deep fake videos.
Violations of the law would result in penalties of up to $200,000 or jail time.
It also restricts the state’s ability to use AI for social scoring, biometric data, or photos of a person without their consent.
State Sen. Charles Schwertner (R-Georgetown) said TRAIGA sets clear guardrails to prevent misuse while allowing innovation to thrive.
“While this evolving technology has the ability to reshape how we live and interact with the world, if left unaddressed, it can bring serious risk of harm to consumers, and an outcomes-based approach to the regulation of AI offers an approach focusing on clearly defined harms that also affords companies the flexibility to innovate around those concerns,” Schwertner said. “This bill establishes a comprehensive framework to ensure AI is developed and deployed responsibly in Texas.”
Big Tech companies have lobbied the federal government for years to override state AI regulations, arguing that the burden of complying with multiple state regulations hinders innovation.
“If there isn’t some sort of institutional body that is helping us decide and creating enforceable guardrails around what appropriate AI use looks like, then the imperative really just pushes us towards greater profit,” Shorey said.
In recent years, as AI has begun to permeate everyday life, states have sought to rein in the rapidly expanding technology. It is increasingly used to make decisions, such as which candidates to interview for a job, home loans, and, in some cases, medical care.
“We don’t know about the safety of this technology, especially when it comes to young people interacting with it. We don’t know if we can trust this technology to make decisions about our health. We don’t know if this technology is going to consolidate data to the point where governments and corporations can surveil American citizens,” Shorey said. “In a lot of ways, the states have been little laboratories for figuring out what this can look like? In what ways do we encourage businesses and encourage development and progress while also protecting basic human rights of the people who live here?”
Texas has also passed legislation requiring the disclosure of AI use in healthcare decisions and limiting AI use in insurance claim determinations.
“AI is not a perfect technology, and when it makes mistakes and denies medical claims, for example, what ends up happening is that it costs Americans everything. When we think about not just access to care, but also the time we spend on the phone and the effort it takes to get through layers of bureaucracy to get a decision reversed, it costs the company nothing. They just decided for free,” Shorey said. “In terms of the time, energy, and administrative burden that falls on people, if the software is running very well, that’s really significant. So by requiring that a human stays in the loop, it allows systems to become more efficient, but it also ensures that there is continued accuracy.”
Without regulations, Shorey said, there could be more instances where AI makes mistakes.
Research has shown cases where it prioritizes a particular gender or race.
“When we talk about AI working imperfectly, it’s not just about improving the technology so that it is smooth and seamless, but it’s actually putting benchmarks in place that help us say this is what it means for a technology to work well,” Shorey said.
The executive order directs the White House to work with Congress to create a national AI framework.
But Shorley said she worries that, in the short term, it could all lead to an environment with little accountability.
“The last thing that we should be doing is creating a void, even if the long-term intention is to create comprehensive legislation in the in between,” Shorey said. “Especially at the rapid pace of development we currently are seeing from AI, we just can’t afford to have an open space that does not take into consideration the real impacts that can happen between the short term and long term.”
Over the summer, Congress killed a Republican attempt, led by Texas U.S. Sen. Ted Cruz, to prevent states from regulating AI. The Senate also voted nearly unanimously to remove a 10-year moratorium on the enforcement of state AI regulation from Trump’s sweeping domestic policy bill, the “Big Beautiful Bill.”
Lawmakers also declined to add an AI moratorium to the National Defense Authorization Act.