HB 149 explained: What Texas’s new AI law (TRAIGA) bans, and the fines behind it
TL;DR
HB 149, the Texas Responsible Artificial Intelligence Governance Act, took effect January 1, 2026. It bars developing or deploying AI meant to push people toward self-harm, violence or crime, or to unlawfully discriminate, and limits how government can use AI. Only the Attorney General can enforce it. Violators get notice and 60 days to fix the problem before a lawsuit, and penalties run up to $200,000 per uncurable violation.
What the bill does
Prohibited uses, for anyone doing business in Texas. The law says a person may not develop or deploy an AI system that intentionally aims to encourage someone to harm themselves (including suicide), harm another person, or commit a crime. It also bars AI built with the intent to unlawfully discriminate against a protected class. The text notes that a “disparate impact” by itself isn’t enough to show intent.
Sexual content involving minors and illegal deepfakes. It prohibits AI developed or distributed with the sole intent of producing child sexual abuse material or illegal sexual deepfakes, as defined in the Penal Code. It also bars text chatbots that imitate a child under 18 in sexual conversations.
Limits on government. State and local government entities can’t use AI for “social scoring,” meaning rating people on their behavior or personal traits in a way that leads to unfair treatment or violates their rights. They also can’t use AI to identify a specific person from biometric data gathered without consent when doing so would violate that person’s rights. A government agency whose AI system talks with the public has to tell people they’re dealing with AI.
Enforcement. The Attorney General has exclusive authority to enforce the law, and it creates no private right to sue. Before filing suit, the AG has to send notice, and the AG can’t sue if the problem is fixed within 60 days.
| Violation type | Civil penalty |
|---|---|
| Curable (or breach of a statement to the AG) | $10,000 – $12,000 each |
| Uncurable | $80,000 – $200,000 each |
| Continuing | $2,000 – $40,000 per day |
Also in the bill: a regulatory “sandbox” run by the Department of Information Resources, where approved participants can test AI systems for up to 36 months, and a new Texas Artificial Intelligence Council.
Authors & sponsors
Author: Rep. Giovanni Capriglione (R), HD 98
Joint authors: Rep. Angie Chen Button (R), Rep. Greg Bonnen (R), Rep. Angelia Orr (R), Rep. Salman Bhojani (D)
Coauthors: Rep. Cody Harris (R), Rep. Suleman Lalani (D), Rep. John Lujan (R), Rep. Will Metcalf (R), Rep. Mihaela Plesa (D)
Senate sponsor: Sen. Charles Schwertner (R), SD 5
Senate cosponsor: Sen. Royce West (D)
Votes
| Vote | Date | Yeas–Nays |
|---|---|---|
| House (passed) | Apr 23, 2025 | 146–3 |
| Senate (passed, amended) | May 23, 2025 | 31–0 |
| House concurs in Senate changes | May 30, 2025 | 121–17 |
Timeline
| Mar 14, 2025 | Filed in the House |
| Apr 23, 2025 | Passed the House |
| May 23, 2025 | Passed the Senate with amendments |
| May 30, 2025 | House agreed to Senate changes |
| Jun 22, 2025 | Signed by the Governor |
| Jan 1, 2026 | Took effect |
Status & what to watch
Status: In effect since January 1, 2026.
What to watch: the Attorney General’s first enforcement actions under the law, and what the sandbox program and AI Council produce.
Sources
- HB 149 bill history (Texas Legislature Online)
- HB 149 enrolled text, with certified vote counts
- HB 149 authors
- HB 149 sponsors
- Party affiliation, 89th Legislature (Legislative Reference Library)