Texas AI Governance — Trustible
Texas Responsible AI Governance Act (TRAIGA) · HB 149 · Tex. Bus. & Com. Code Ch. 552

Texas AI Governance

Texas took a narrower path than Colorado. The Texas Responsible AI Governance Act (TRAIGA) is not a universal high-risk-AI framework with mandatory impact assessments. It is a targeted set of prohibitions enforced exclusively by the Texas Attorney General, paired with sector guidance from the Texas Department of Insurance that reaches AI-assisted decisions across every regulated carrier in the state. Organizations selling into Texas or regulated by TDI need both pieces in view.

Requirement
What It Means
Consumer AI Disclosure
Applies to government agencies generally, and separately to healthcare providers using AI in treatment, who must disclose AI use to the patient by the time service is first provided. There is no blanket private-sector AI disclosure duty.
No AI-Driven Manipulation
Bans intentionally developing or deploying AI to incite self-harm, suicide, harm to others, or criminal activity. The standard is intent-based, so it is narrow in practice.
Government Social Scoring Ban
Restricts government entities only from using AI to score or classify people by social behavior in ways that cause unjustified detriment or infringe legal rights.
Government Biometric Capture Ban
Restricts government entities only from using AI plus biometric or internet-scraped image data to identify individuals without consent in a way that infringes constitutional or legal rights. HIPAA-covered data is excluded.
Unlawful Discrimination
Bans developing or deploying AI with intent to discriminate against a protected class. Disparate impact alone is explicitly not enough to establish intent, a key narrowing amendment from the original bill. Insurance entities and federally regulated financial institutions have separate carve-outs.
CSAM and Deepfake Prohibition
Bans AI developed with the sole intent to produce CSAM or unlawful deepfakes, including chatbots that simulate sexual conversations while impersonating a minor.

Enforcement runs exclusively through the Texas Attorney General. There is no private right of action. The AG can issue civil investigative demands covering a system’s purpose, training data, inputs and outputs, performance metrics, known limitations, and monitoring documentation, and must give 60 days’ notice and an opportunity to cure before filing suit. Civil penalties range from $10,000 to $12,000 per curable violation, $80,000 to $200,000 per uncurable violation, and $2,000 to $40,000 per day for continuing violations. TRAIGA also creates an affirmative defense for organizations that can show substantial compliance with NIST’s AI RMF or a comparable recognized risk-management framework, and a DIR-administered regulatory sandbox lets approved participants test AI systems for up to 36 months without standard state licensing (the core prohibitions still apply inside the sandbox). Chapter 552 also preempts city and county AI ordinances statewide, which matters for organizations operating across multiple Texas municipalities.

Requirement
What It Means
Governance and Risk-Management Controls
Requires documented controls over AI development, acquisition, and use across regulated insurance operations.
Human Review
Requires human review of consequential AI-supported decisions in underwriting, pricing, and claims.
Bias and Error Testing
Requires bias and error testing before and during deployment of AI systems.
Vendor Accountability
Using a third-party model does not shift liability away from the carrier — accountability for vendor AI stays with the insurer.

Enforced by the Texas Department of Insurance through its examination authority over authorized carriers, HMOs, and utilization review agents operating in the state.

How Trustible Supports Texas Compliance

Compliance
Trustible Capability
AI Inventory
Centralizes every AI system, developer relationship, and deployment context so TRAIGA’s developer/deployer distinctions are documented, not inferred after the fact.
Automated Workflows
Routes AI use cases through structured intake and review so consumer disclosure obligations, healthcare treatment disclosures, and vendor oversight steps are never skipped.
Risk Management
Documents intent-based discrimination review with clear rationale and evidence, supporting TRAIGA’s affirmative defense standard and TDI’s bias-testing expectations.
Insights Taxonomies
Expert-curated risk categories keep discrimination, manipulation, and biometric risk definitions current as TRAIGA enforcement guidance develops.
Model and Vendor Evaluations
AI-assisted analysis of vendor documentation surfaces the transparency gaps TDI expects carriers to identify before accountability shifts back to them.
Reporting & Dashboards
Generates the audit-ready documentation the Texas AG can request through a civil investigative demand, before a demand ever arrives.

Your First 90 Days

Day 30: Map Developer and Deployer Roles

Identify every AI system in use, who developed it, and who deploys it in Texas. Flag any healthcare, insurance, or government-adjacent use cases subject to heightened disclosure duties.

Day 60: Document Discrimination and Bias Review

Launch structured review of AI systems for discriminatory outcomes, with documented rationale that goes beyond disparate impact data alone. For insurance use cases, align testing with TDI Bulletin B-0003-26 expectations.

Day 90: Operationalize Vendor Oversight and Reporting

Connect vendor evaluation workflows to procurement so third-party AI accountability doesn’t quietly transfer to Trustible customers. Prepare documentation TRAIGA’s cure period and TDI examinations would require on short notice.

Texas AI FAQs

No. The general AI disclosure duty applies to government agencies. Private companies only face a specific disclosure duty in the healthcare treatment context.

No. TRAIGA requires intent to discriminate. A narrowing amendment made clear that disparate impact data by itself is not sufficient to establish that intent, a meaningfully different standard than Illinois or Colorado.

The Texas Attorney General has exclusive enforcement authority. There is no private right of action, and the AG must provide 60 days’ notice and a cure opportunity before suing.

Yes. Chapter 552 preempts city and county regulation of AI systems, so compliance obligations are set at the state level statewide.

TDI Bulletin B-0003-26 applies existing insurance law to AI-assisted decisions for every authorized carrier, with no size threshold, requiring governance controls, human review, and vendor accountability independent of TRAIGA’s own provisions.

See How Trustible Operationalizes Texas AI Compliance in a Unified Governance Program.

© 2026 Trustible