Texas Responsible Artificial Intelligence Governance Act (TRAIGA)

Summary

TRAIGA stands for the Texas Responsible Artificial Intelligence Governance Act, the state's first comprehensive AI statute. Enacted as House Bill 149 in the 89th Texas Legislature and signed June 22, 2025, TRAIGA is codified at Texas Business and Commerce Code Chapters 551 through 554 and took effect January 1, 2026. The Texas Legislature's current bill history and codified Business and Commerce Code still show that effective date and chapter structure; the § 552.105 civil-penalty tiers and § 552.106 licensing-sanction cap were re-verified against the enrolled HB 149 text on 2026-08-08. The Act applies to persons doing business in Texas who develop or deploy artificial intelligence systems and creates obligations targeted at AI consumer disclosures, manipulation that incites self-harm or violence, government social scoring, biometric capture, unlawful discrimination, and certain sexually explicit synthetic content.

Key provisions include clear and conspicuous disclosure when covered AI systems interact with consumers, prohibitions on AI systems intentionally developed or deployed for unlawful discrimination, restrictions on government use of AI for social scoring, a complaint intake mechanism at the Texas Attorney General's office, a regulatory sandbox program, and the Texas Artificial Intelligence Council. Related Atlas coverage: Utah AI Policy Act.

Civil-penalty tiers at a glance

• Curable violation, not cured — $10,000 to $12,000 per violation (Tex. Bus. & Com. Code § 552.105).
• Uncurable violation — $80,000 to $200,000 per violation (§ 552.105).
• Continuing violation — $2,000 to $40,000 per day until corrected (§ 552.105).
• State licensing-agency sanctions — up to $100,000 against licensed, registered, or certified persons (§ 552.106), separate from AG enforcement.

Enforcement and notice-and-cure

The Texas Attorney General has exclusive enforcement authority under Tex. Bus. & Com. Code § 552.101; TRAIGA does not create a private right of action. Before suit, the AG issues written notice under § 552.104, giving the entity an opportunity to cure curable violations and avoid the tier-1 civil penalty. Failure to cure, or an uncurable violation (for example, intentional unlawful discrimination prohibited under § 552.056), authorizes the tier-2 and continuing-violation penalties under § 552.105.

Sector implementation routes

Because TRAIGA applies by Texas business activity and AI-system use rather than a single regulated vertical, compliance teams usually need to map the same Chapter 552 controls into sector workflows. Start with financial services AI compliance for credit, lending, fraud, and account-servicing AI; healthcare AI compliance for patient-facing and regulated-occupation AI interactions; HR and hiring AI compliance for employment decisions; AI in insurance compliance for underwriting, claims, and pricing; and the education, housing, legal services, and government services hubs when Texas deployments touch those operating areas.

Sandbox and Texas Artificial Intelligence Council

TRAIGA establishes a regulatory sandbox program and the Texas Artificial Intelligence Council to advise the legislature on emerging AI policy questions; both sit alongside the Chapter 552 substantive prohibitions. Companies considering Texas-only AI pilots should review sandbox eligibility before deploying covered systems against consumers and map the resulting control owners in an AI governance model and AI compliance framework register.