“Texas lawmakers push to regulate AI in government and the tech industry” was first published by The Texas Tribune, a nonprofit, nonpartisan media organization that informs Texans — and engages with them — about public policy, politics, government and statewide issues.

Sign up for The Brief, The Texas Tribune’s daily newsletter that keeps readers up to speed on the most essential Texas news.

With AI tools embedded in everyday life — from customer service chatbots and ChatGPT to predictive policing algorithms — Texas is seeking to place boundaries around the fast-growing technology by imposing a host of rules and appointing “a new sheriff in Texas’ digital town.”

House Bill 149, authored by state Rep. Giovanni Capriglione, R-Southlake, is Texas’ attempt to create guardrails that allow innovation while protecting people from potential harm, said state Sen. Charles Schwertner, R-Georgetown, the bill’s Senate sponsor, at a recent committee hearing.

The bill would require government agencies to disclose when Texans are interacting with an AI system on a state agency website and ban the capture of biometric identifiers without consent — including retina, iris or facial scanning, fingerprints and voice prints. The bill also would prohibit industry from developing AI systems designed to manipulate human behavior and prohibit discrimination and deep fake child exploitation.

The Texas Attorney General’s office would be charged with enforcing the bill, aided by an online complaint system. Violators would face civil fines of up to $100,000.

“I don’t think yet we really need to worry about a Terminator scenario of killer robots,” said Kevin Welch, president of EFF-Austin, a consumer advocacy group that advocates for the protection of digital rights. “I would say it’s important to focus on real harms, which is one thing I do really like about this bill. It focuses on real harms and not hypothetical sci-fi scenarios.”

Supporters say the bill is a necessary first step to prevent harms like racial profiling, privacy violations, or opaque government decision-making. Critics have warned that the bill could stifle innovation and may introduce legal uncertainty if not tweaked to clarify certain language.

David Dunmoyer, the campaign director for Better Tech for Tomorrow at the Texas Public Policy Foundation, a nonprofit research institute based in Austin, says the bill is about “getting AI policy right before the whole horse is out of the barn,” which means getting the “right guardrails and the right regulatory system in place that ensures we’re not just preserving humanity, but advancing it and furthering it.”

He said the bill focuses on outcomes by drawing clear boundaries around what AI should not be allowed to do and increasing transparency.

“Really it boils down to balancing the need for some policy and regulation around this to protect people’s privacy and their transparency and the need to not stifle innovation,” said Sherri Greenberg, an AI expert and assistant dean for state and local government engagement at the UT-Austin. Greenberg added that the attorney general would have the authority to enforce AI regulations regardless of where the AI system is based.

The bill would prohibit government agencies from using AI systems to assign “social scores” or rank people based on personal data. In the private sector, developers would be prohibited from designing AI tools that incite self-harm, violence, or criminal behavior. The bill also would restrict the use of AI to limit a person’s access to political content or infringe on freedom of expression or association.

The bill would also create the Texas Artificial Intelligence Council, housed within the state Department of Information Resources. The 10-member advisory body would monitor AI use across state government, flag harmful practices, recommend legislative updates and identify rules that may be impacting innovation.

For the AI industry, the bill creates a regulatory “sandbox,” a controlled environment where developers can test AI systems free from certain state rules without being penalized. Lawmakers have said the sandbox is designed to balance technological freedom with public oversight.

The bill has been approved by both chambers and is headed back to the House, where representatives will decide if they agree with amendments added by the Senate or need to go to a conference committee to hash out the differences.

If approved, the bill would come with a $25 million price tag and add 20 new full-time staff positions, including 12 in the AG’s office.

Even if the law passes, its impact could be short-lived if Congress steps in. A recent draft of the 2025 federal budget reconciliation bill would put a 10-year moratorium on new state AI laws, which could freeze bills like HB 149 before they take effect. HB 149, if it becomes law, would take effect on Jan. 1.

Would the bill protect citizens harmed by AI?

Dunmoyer, who testified in support of the bill at a recent hearing, said that the bill addresses industry’s concern of getting punished for trying to innovate.

“This bill seeks an environment of compliance rather than punishment,” he said, adding that the bill “provides what the industry has asked for, which is clear rules of the road and protection against a litigious hellscape.”

While the bill offers some flexibility for industry to address potential harms, Welch, president of the consumer advocacy group EFF-Austin, said the bill prohibits private right of action, meaning it blocks citizens from suing companies that violate their rights through AI.

“I feel like [these laws] often end up as being a lot of nice words and sentiments, but the actual rights of citizens aren’t protected,” he said. “I do feel that if we really want to give these laws teeth, we have to make it where citizens can bring lawsuits.”

Dunmoyer said that the bill creates a new online portal where Texans can submit complaints to the attorney general, who he calls “a new sheriff in Texas’ digital town,” to investigate potential violations.

Meanwhile, Anton Dahbura, co-director of the Johns Hopkins Institute for Assured Autonomy, said regulating AI is far more complicated than lawmakers seem to realize. He argues that AI isn’t a single, well-defined entity, it’s really a broad and evolving field of technologies and techniques.

This misunderstanding, he warns, leads to misguided attempts at regulation that may not be enforceable or effective.

Dahbura remains neutral on whether AI should be regulated, but he stresses that any such efforts need to be informed and precise. He suggests that regulation should focus on outcomes — holding people accountable for harm or illegal actions regardless of the tools used — instead of trying to legislate the technology itself.

Dahbura said he also sees a problematic narrative forming around AI as a threat that must be neutralized, likening it to a “pitchforks and torches” approach.

“It feels a little bit like people are marching up the hill to get the bad guy that is AI,” he said. “And if they corner the bad guy, then everything is great.”

The risk of taking regulations too far, he added, is placing unnecessary and ill-conceived burdens on the industry, potentially stifling innovation without offering real protections for people.

Where AI regulation started

Lawmakers have created other advisory bodies aimed at studying the impacts of AI. In 2023, the state approved a bill that created an Artificial Intelligence Advisory Council within the Department of Information Resources to study how AI systems are used in state government, whether they infringe on legal or constitutional rights and recommending ethical guidelines. That council was disbanded after submitting its report to lawmakers in December 2024.

A separate AI and Emerging Technologies Select Committee also made recommendations such as requiring state agencies to audit their AI systems annually, provide state employees with training on AI ethics and data privacy and the formation of an AI sandbox. This recommendation led to HB 149’s sandbox program.

A major concern raised during initial hearings was the deceptive potential of AI — from cloned voices to deepfakes — and how such technologies could undermine democracy and public trust. By mid-2024, agencies were required to report their AI activities to the advisory council, whose findings informed this year’s legislation.

Capriglione, who also championed HB 4, the state’s landmark data privacy law, played a central role. Alongside Sen. Tan Parker, R-Flower Mound, they held meetings with experts in AI, consumer advocacy and technology to figure out what responsible AI governance should look like in Texas. Out of these meetings came a bill by Parker focused on regulating AI within government agencies and Capriglione’s first attempt to regulate AI in the private sector.

Capriglione’s original proposal, HB 1709 — also known as the Texas Responsible AI Governance Act — focused on regulating AI in health care, employment, and finance. It was modeled after the European Union’s AI Act. But the tech industry pushed back, calling it too broad and burdensome.

“Regulating AI in industry is a more difficult proposition,” said Greenberg, the UT-Austin AI expert. “You may get pushback from industry saying that this is going to put us behind or stifle innovation.”

The bill never made it to a House committee. Capriglione came back with HB 149.

Across the country, nearly every state in the country introduced legislation related to AI this year, while others already have laws on the books.

Texas carefully studied Colorado’s AI law, which was signed into law this month and targets AI systems used in decisions related to education, employment, financial services, government services, health care, housing, insurance, or legal services. The bill aims to prevent discrimination based on protected characteristics like age, race or gender.

​​Texas lawmakers are also considering other AI-related bills during this legislative session, which ends June 2. One would require that political advertisements disclose whether images, audio, or video have been substantially altered using AI. Another bill seeks to prohibit AI-generated child pornography.

Disclosure: The Texas Public Policy Foundation has been a financial supporter of The Texas Tribune, a nonprofit, nonpartisan news organization that is funded in part by donations from members, foundations and corporate sponsors. Financial supporters play no role in the Tribune’s journalism. Find a complete list of them here.

First round of TribFest speakers announced! Pulitzer Prize-winning columnist Maureen Dowd; U.S. Rep. Tony Gonzales, R-San Antonio; Fort Worth Mayor Mattie Parker; U.S. Sen. Adam Schiff, D-California; and U.S. Rep. Jasmine Crockett, D-Dallas are taking the stage Nov. 13–15 in Austin. Get your tickets today!

This article originally appeared in The Texas Tribune at https://www.texastribune.org/2025/05/23/texas-ai-bill-legislation-regulation/.

The Texas Tribune is a member-supported, nonpartisan newsroom informing and engaging Texans on state politics and policy. Learn more at texastribune.org.

The post Texas bill seeks to regulate AI’s use in the state appeared first on feeds.texastribune.org

Note: The following A.I. based commentary is not part of the original article, reproduced above, but is offered in the hopes that it will promote greater media literacy and critical thinking, by making any potential bias more visible to the reader –Staff Editor.

Political Bias Rating: Center-Right

This article presents Texas’ AI regulation bill largely through a lens favorable to Republican lawmakers and conservative think tanks, emphasizing balanced regulation that protects privacy and innovation without overburdening industry. It highlights sponsors from the GOP and includes supportive quotes from groups like the Texas Public Policy Foundation. The tone remains measured, acknowledging critics and nuanced perspectives but framing the legislation as a pragmatic, innovation-friendly approach. While it reports on concerns from consumer advocates and academics, the overall framing leans toward a center-right viewpoint that values market-driven solutions and limited government intervention balanced with necessary safeguards.