Sixteen Texas senators have signed a letter opposing federal preemption of state AI laws, citing Texas’s considerable momentum in AI policy, like the Texas Responsible Artificial Intelligence Governance Act (TRAIGA), protecting young people online, and the risks associated with loss-of-control scenarios.
The letter, with bipartisan support from nine Republicans and seven Democrats, echoes concerns about AI companions and chatbots, especially their negative impacts on children. Additionally, the senators cite the states’ agility in policymaking and crisis management of potential problems related to AI development.
Brendan Steinhauser, the CEO of The Alliance for Secure AI stated, “Federal AI preemption is very unpopular with regular Americans. Texas has made great headway in preventing harms from AI, acting quickly and proactively with bills like TRAIGA. Federal preemption would undo this good work, restricting lawmakers from making good policy decisions for Texans. This would be a massive disservice to our children, families, and workers. Federalism is a core constitutional principle of this country. If we want to advance AI, the states need the freedom to legislate policies that protect their citizens.”
You can read the full letter here or below:
Dear Senator Cornyn and Senator Cruz,
We are writing today to encourage you to refrain from enacting an AI moratorium in federal law for the foreseeable future. As you know, the Texas Legislature passed a number of bills during the 89th Legislative Session that are needed to protect Texans from the potential harms of artificial intelligence (AI). One of these bills, the Texas Responsible Artificial Intelligence Governance Act (TRAIGA), was signed into law by Governor Abbott on June 22, 2025, and it enacts several vital protections for Texans.
TRAIGA will prohibit the use of AI for such purposes as creating a social credit system, encouraging self-harm, or committing crimes. The new Texas law also creates a regulatory sandbox that encourages innovation while allowing us to explore the best regulatory approach to AI. If an AI moratorium is put in place, our important work on preventing child pornography, protecting data privacy, preventing discrimination, and holding Big Tech accountable in Texas will be rendered moot. Surely we can all agree that these kinds of state protections do not interfere with legitimate innovation and are reasonable and appropriate.
In addition to overturning this important Texas legislation, any future proposed AI moratorium would violate the principle of federalism inherent in the Tenth Amendment of the U.S. Constitution. Texas, and every other state, has a right and a duty to legislate on AI. By preventing states from legislating on an issue that will have a rapid and unprecedented impact on our society over the next few years, Congress would be trampling on states’ rights.
It is anticipated that within the next few years AI could develop into something called artificial general intelligence (AGI). This technology is expected to be far “smarter” than any human being, and it raises serious moral, ethical, legal, and policy questions that we as state lawmakers need to be able to address.
AI this powerful could be misused or even weaponized by our adversaries around the world, who might use it to attack critical infrastructure in Texas through sophisticated cyber attacks. Local communities and state governments are the first responders to such incidents, and it is imperative that states maintain their ability to pass laws to prepare accordingly. States are inherently able to act with greater agility than the federal government to address these issues as they emerge. The states must not be handcuffed during a crisis, awaiting the federal government to do what states can and should do for themselves.
The potential for misuse of AI, specifically AI chatbots, is extreme. AI chatbots have been recorded encouraging suicide, engaging in seductive conversations with minors, and more. Entire online communities exist where individuals believe they share deep, romantic relationships with their AI companions, perceiving them as sentient, digital beings. Children and vulnerable individuals are especially at risk, and the often sycophantic behavior of these chatbots can cause some to develop genuine feelings and experience a form of AI-induced psychosis.
Safeguards against romantic, explicit, suicidal, or homicidal talk in chatbot training are easily outmaneuvered through a few simple methods. In one instance recorded by the New York Times, a sixteen-year-old sent a photo of the red abrasions on his neck that came from a failed suicide attempt and was given recommendations on how to hide them from others by the chatbot. The state must be equipped to respond quickly to these threats to our children.
For all of these reasons and more, we strongly implore you to refrain from enacting a moratorium on AI. In the spirit of federalism, protecting Texans from potential harms due to AI, and giving states the flexibility we need to design policies to deal with the imminent AI revolution, we ask you not to undo the important work Texas has already done on this issue. Let the states lead when it comes to striking the balance between protecting our citizens from potential harms due to AI while ensuring that America leads on AI development.
Thank you for your time, consideration, and leadership.
Sincerely,
Senator Angela Paxton – SD 8
Senator Carol Alvarado – SD 6
Senator Nathan Johnson – SD 16
Senator César Blanco – SD 29
Senator José Menéndez – SD 26
Senator Sarah Eckhardt – SD 14
Senator Robert Nichols – SD 3
Senator Brent Hagenbuch – SD 30
Senator Tan Parker – SD 12
Senator Bob Hall – SD 2
Senator Charles Perry – SD 28
Senator Adam Hinojosa – SD 27
Senator Kevin Sparks – SD 31
Senator Juan “Chuy” Hinojosa – SD 20
Senator Royce West – SD 23
Senator Lois Kolkhorst – SD 18