Texas Responsible Artificial Intelligence Governance Act

Quick Overview

Texas House Bill 149, the Responsible Artificial Intelligence Governance Act, passed the legislature and is set to become effective on January 1st, 2026, establishing a regulatory framework that emphasizes transparency, requires AI developers and deployers to explain model outputs, and creates a specialized AI Council to guide future laws, while imposing significant financial penalties for non-compliance, especially for uncured violations.

Key Points: Texas House Bill 149, the Responsible Artificial Intelligence Governance Act, officially cites as Section 551.001 through 551.108, becomes effective on January 1st, 2026. The bill mandates transparency, requiring government agencies using AI to disclose when a resident is interacting with an AI system, not a person. It establishes an AI Council composed of 10 members appointed by the Governor, Lieutenant Governor, and Speaker, requiring expertise in AI, data privacy, ethics, and antitrust law. The Council's role is advisory, meant to guide future laws, not to create binding rules, though they must report findings back to the legislature. The bill bans government entities from using AI to create social scoring or to discriminate based on political viewpoints, and prohibits using AI to generate sexually explicit deep fakes. Non-compliance with disclosure rules for high-risk systems can result in daily fines between $2,000 and $40,000 per violation. The bill offers a 36-month sandbox program for companies to test new AI systems without meeting full regulatory compliance immediately.

Context: The video discusses the implications of the recently passed Texas Responsible Artificial Intelligence Governance Act (HB 149), which represents a significant shift in the regulatory landscape for AI deployment within the state. The act aims to balance fostering AI innovation with protecting citizens' rights and ensuring governmental accountability regarding automated decision-making systems.

Detailed Analysis

Texas House Bill 149, the Responsible Artificial Intelligence Governance Act, is a major piece of legislation set to take effect on January 1, 2026. The bill significantly broadens the regulatory scope beyond traditional deterministic software to cover probabilistic systems, which the text defines as AI systems that use data to train statistical models. A core requirement is transparency, mandating that any governmental agency using AI to communicate with a Texas resident must clearly disclose that the interaction is with an AI, not a human. Furthermore, the bill establishes an AI Council, a 10-member body appointed by the Governor, Lieutenant Governor, and Speaker, requiring members with expertise in AI, data privacy, ethics, and antitrust law. This Council serves an advisory role to guide future legislation, not as a rulemaking body. The bill explicitly bans certain uses, such as employing AI for social scoring or discrimination based on political viewpoints, and prohibits the creation of sexually explicit deep fakes. Non-compliance is penalized heavily, with daily fines ranging from $2,000 to $40,000 for violations like failing to disclose the use of AI or using AI to manipulate users. However, the bill includes a 36-month sandbox program allowing companies to test new AI systems without immediate full regulatory compliance, provided they submit a detailed plan outlining benefits and risk mitigation. The Attorney General has the authority to enforce these rules, including seeking civil investigative demands and potentially levying fines, although the bill grants a right to cure for non-willful violations.