Sign up for Brief, the Texas Tribune’s daily newsletter, to keep your readers in the most important Texas news.
With AI tools embedded in everyday life, from customer service chatbots and chatbots to predictive policing algorithms, Texas aims to place boundaries around the fast-growing technology by imposing a number of rules and appointing a “new sheriff in Texas’ digital town.”
Written by R-Southlake State Sen. Giovanni Capriglione, House Bill, at a recent committee hearing, state Sen. Bill, state Sen. Charles Schwartner, is a Texas attempt to create guardrails that allow innovation while protecting people from potential harm.
The bill requires government agencies to disclose if Texans interact with AI systems on state agency websites and prohibit the capture of biometric identifiers without consent, such as retina, iris, facial scans, fingerprints, and audio prints. The bill also prohibits the industry from developing AI systems designed to manipulate human behavior and prohibit discrimination and deep false child exploitation.
The Texas Attorney General’s Office will be charged with enforcement of the bill, with the support of the online complaint system. Violators will face civil fines of up to $100,000.
“I don’t think we need to worry about the killer robot terminator scenario yet,” said Kevin Welch, president of Eff-Austin, a consumer advocacy group that advocates for digital rights protection. “I think it’s important to focus on real harm. This is one of the things I really like about this bill. It focuses on real harm, not hypothetical science fiction scenarios.”
Supporters say the bill is the first step necessary to prevent harm from racial profiling, privacy violations, or opaque government decision-making. Critics warn that if the bill is not tweaked to curb innovation and clarify a particular language, it could introduce legal uncertainty.
David Dunmoyer, campaign director for Tomorrow’s Technology at the Texas Public Policy Foundation, an Austin-based non-profit laboratory, said the bill was “about getting an AI policy just before the whole horse leaves the barn.”
He said the bill focuses on outcomes by clearly leading to the boundaries that AI should not be allowed and increasing transparency.
“We are pleased to announce that we are committed to providing AI expert and assistant dean for UT-Austin’s state and local government engagement. Greenberg added that the Attorney General has the authority to implement AI regulations regardless of the foundations of the AI system.
The bill would prohibit government agencies from using AI systems to assign “social scores” or ranking people based on personal data. In the private sector, developers are prohibited from designing AI tools that incite self-harm, violence, or criminal behavior. The bill also restricts the use of AI, restricting people’s access to political content and infringes freedom of expression and associations.
The bill also establishes the Texas Artificial Intelligence Council, housed within the state Department of Information and Resources. The 10 advisory bodies monitor AI use across state governments, flag harmful practices, recommend legislative updates, and identify rules that could impact innovation.
For the AI industry, the bill creates a regulatory “sandbox.” This is a regulatory environment where developers can test AI systems free of charge from certain state rules without being punished. Lawmakers say the sandbox is designed to balance technology freedom with public surveillance.
The bill has been approved by both rooms and will return to the House of Representatives. There, the representative decides whether to agree to the amendments that the Senate has added.
If approved, the bill will have a price tag of $25 million and add 20 new full-time staff positions, including 12, to AG’s office.
Even if the law is passed, if Congress intervenes, the impact could be short-lived. The recent draft of the 2025 Federal Budget Settlement Bill places a 10-year suspension on new state AI laws. HB 149 will take effect on January 1st if it becomes law.
Will the bill protect citizens hurt by AI?
Testifying in support of the bill at a recent hearing, Dunmoyer said the bill is addressing industry concerns that it will be punished for attempting to innovate.
“The bill calls for an environment of compliance rather than a punishment,” he said, adding that the bill “provides what the industry is looking for: it’s clear rules on the roads and protection against the hellish scenery of litigation.”
While the bill offers some flexibility for the industry to address potential harms, Welch, president of consumer advocacy group Eff-Austin, said the bill prohibits the right to private conduct.
“I think (these laws) are often a lot of nice words and feelings, but the actual rights of citizens are not protected,” he said. “If we really want to give these laws teeth, we feel we have to make them in a place where citizens can file lawsuits.”
Dunmoyer said the bill would create a new online portal where the Texans can file complaints with the Attorney General, whom they call “the new sheriff in Texas digital towns” to investigate potential violations.
Meanwhile, Anton Darbra, co-director of the Johns Hopkins Institute of Assurance and Autonomy, said AI regulations are far more complicated than lawmakers seem to recognize. He argues that AI is not a single, well-defined entity, but is actually a broad and evolving field of technology and techniques.
This misconception warns it is a warning, leading to false attempts at regulation.
Dahbura remains neutral on whether AI should be regulated, but emphasizes that such efforts need to be informed and accurate. He suggests that regulations should not attempt to legislate the technology itself, but rather focus on the consequences of holding people accountable for harm and illegal activities, regardless of the tools they use.
Dahbura said that the problematic narratives formed around AI are formed as threats that must be neutralised, and compares them to a “pitchfork and torch” approach.
“It feels like people are marching up the hill to get the bad guys who are AI,” he said. “And if they’re going to corner the bad guys, everything’s great.”
The risk of overregulation is putting unnecessary and inappropriate burden on the industry, he added, which could limit innovation without providing real protection to people.
Where AI regulations began
Lawmakers have created other advisory bodies aimed at studying the impact of AI. In 2023, the state created an AI Advisory Committee for the Information Resources Agency to approve a bill that would study how AI systems are used by state governments, violate legal or constitutional rights, and recommend ethical guidelines. The council was dissolved in December 2024 after submitting a report to lawmakers.
Another AI and Emerging Technologies Select Committee required state agencies to audit their AI systems annually, and created recommendations for state employees, including training on AI ethics and data privacy and forming an AI sandbox. This recommendation led to the HB 149 sandbox program.
The major concerns raised at the first hearing were the deceptive possibilities of AI, from cloned voices to deepfakes, and the way such technology undermines democracy and public trust. By mid-2024, institutions had to report their AI activities to the Advisory Committee.
Capriglione, who defended the state’s groundbreaking data privacy law, HB 4, played a central role. Alongside Senator Tan Parker of R-Flower Mound, they held meetings with AI experts to understand what responsible AI governance looks like in Texas, consumer advocacy and technology. Of these meetings, there was a bill by Parker that focused on Capriglione’s first attempt to regulate AI within government agencies and private sector AI.
Capriglione’s original proposal, HB 1709 (also known as the Texas Liability AI Governance Act), focuses on regulating AI in healthcare, employment and finance. It was modelled on the European Union’s AI law. But the tech industry is pushing back and calling it a broad and burden.
“Regulating AI in the industry is a more challenging proposition,” said Greenberg, an expert at UT-Austin AI. “We might get pushback from the industry saying this will put us behind us or curb innovation.”
The bill did not reach a House committee. Capriglione returned on HB 149.
Nearly every state across the country has introduced AI-related laws this year, while other states already have book laws.
Texas has carefully studied Colorado’s laws that were signed into law this month and targeted AI systems used to determine related to education, employment, financial services, government services, healthcare, housing, insurance, or legal services. The bill aims to prevent discrimination based on protected characteristics such as age, race, and gender.
Texas lawmakers are also considering other AI-related bills during the legislative session, which ends June 2nd. Political ads must use AI to disclose whether images, audio, or video have been significantly altered. Another bill seeks to ban AI-generated child pornography.
Disclosure: The Texas Public Policy Foundation is a financial advocate for the Texas Tribune, a non-profit, nonpartisan news organization that is funded in part by donations from members, foundations and corporate sponsors. Financial supporters have no role in Tribune journalism. Find the complete list of them here.
The first round of Tribfest speakers has been announced! Pulitzer Prize-winning columnist Maureen Dowd. US lawmakers Tony Gonzalez, R-San Antonio; Fort Worth Mayor Matty Parker; US Senator Adam Schiff, D-California. Rep. Jasmine Crockett of D-Dallas will be on stage in Austin from November 13th to 15th. Get your tickets today!
