Texas Signs Responsible AI Governance Act Into Law

A client alert reports that Governor Greg Abbott of Texas put his signature on the state's Responsible AI Governance Act, known as TRAIGA, on June 22, 2025. The original draft, introduced in December 2024, followed the Colorado and EU models and centered on high-risk systems. An amended version in March 2025 scaled it back, deleting or confining to governmental entities duties like impact assessments and consumer disclosures. The law starts on January 1, 2026.

As enacted, it bars the development or deployment of AI systems for certain purposes by private entities active in Texas. The list covers intentionally pushing people toward self-harm, harming others or crime; aiming to infringe federal constitutional rights; intending unlawful discrimination; and producing child sexual content or unlawful deepfakes. A disparate impact alone does not prove intent to discriminate, and banks and insurers receive tailored treatment. Social scoring and biometric identification are banned only for governmental entities.

Enforcement belongs solely to the state Attorney General, who must provide an online complaint channel and may issue civil investigative demands. A party receives 60 days to cure after notice. Fines range from $10,000 to $12,000 for curable violations, $80,000 to $200,000 for uncurable ones, and up to $40,000 daily for continuing ones. Licensing agencies can add sanctions of up to $100,000. Defenses include red-teaming, agency guidance, internal review and alignment with a recognized framework such as NIST's.

The act also creates a sandbox run by the Department of Information Resources, giving accepted participants 36 months and requiring quarterly reports, plus a seven-member advisory council that may not issue binding rules.