EveryAILaw.com

Requirements at a glance

This regulation imposes 11 specific requirements for Bias & Discrimination Prevention across 1 provision:

• Intentional discrimination prohibited — No person may develop or deploy an AI system with **intent** to unlawfully discriminate against a protected class
• Disparate impact insufficient — Disparate impact alone does not establish intent to discriminate (§ 552.056(c))
• Subtitle applicability alternatives — Subtitle D applies only to a person who promotes, advertises, or conducts business in Texas; produces a product or service used by Texas residents; or develops or deploys an AI system in Texas (§ 551.002)
• Insurance-services exception — Section 552.056 does not apply to an insurance entity **for purposes of providing insurance services** if it is subject to the specified laws governing unfair discrimination, competition, or deceptive practices in insurance. “Insurance entity” includes the developer of an artificial intelligence system used by a listed insurer or fraternal benefit society (§ 552.056(a)(2), (d))
• Banking compliance rule — A federally insured financial institution is considered compliant with § 552.056 only if it complies with **all federal and state banking laws and regulations** (§ 552.056(e))
• Exclusive enforcement and no private action — The attorney general has exclusive Chapter 552 enforcement authority except for the qualified state-agency route in § 552.106; Chapter 552 supplies no private right of action (§ 552.101)
• Notice and 60-day cure — The attorney general may not sue before the **60th day** after written notice and may not sue if, before that day, the person actually cures, provides a written statement with supporting documentation showing how it cured, and makes necessary changes to internal policies to reasonably prevent recurrence (§ 552.104)
• Uncured civil penalties — For an uncured violation, § 552.105(a) sets $10,000–$12,000 for each curable violation or breach of a cure statement, $80,000–$200,000 for each uncurable violation, and $2,000–$40,000 for each day a violation continues
• Qualified liability defenses — A defendant may not be found liable under § 552.105(e) for another person’s prohibited use of its affiliated system. The discovery alternatives are separate: the defendant discovers a violation through feedback; through testing, including adversarial or red-team testing; by following applicable state-agency guidelines; **or through an internal review process if the defendant substantially complies** with the current NIST Generative AI Profile or another recognized AI risk-management framework. The framework-compliance condition belongs only to the internal-review route, and it is not blanket immunity
• Undeployed-system civil-penalty limit — The attorney general may not bring an action to collect a § 552.105 civil penalty for an AI system that has not been deployed (§ 552.105(f))
• Qualified state-agency sanctions — A state agency may sanction a person licensed, registered, or certified by that agency only after the person has been found in violation in a § 552.105 court action and the attorney general recommends added agency enforcement. Sanctions may include suspension, probation, or revocation of a license, registration, certificate, or other authorization, plus a monetary penalty not to exceed $100,000 (§ 552.106)