On December 23, 2024, House Bill 1709 for the Texas Responsible Artificial Intelligence Governance Act was filed in the House of Representatives. The bill aims to regulate the use of artificial intelligence (AI) systems by business entities and state agencies and protect consumers against algorithmic discrimination.

Scope of the bill

The bill applies to:

• large businesses and entities operating in Texas that develop, deploy, or distribute high-risk AI systems influencing consequential decisions, such as employment, healthcare, housing, and financial services;
• digital service providers and social media platforms facilitating the use of high-risk AI systems, particularly if they expose users to algorithmic discrimination or prohibited uses; and
• businesses based outside Texas that develop, distribute, or deploy high-risk AI systems used or consumed by Texas residents or operate platforms or services that collect and process data from Texas residents.

However, small businesses are exempt from the bill's requirements unless they engage in activities directly regulated by the bill.

Requirements under the bill

Under the bill deployers of high-risk AI systems are required to, among other things:

• use reasonable care to protect consumers from algorithmic discrimination, which is defined as unlawful bias based on protected classifications;
• conduct annual impact assessments for each deployed high-risk AI system;
• disclose to consumers interacting with the AI system, that they are interacting with an AI system, the system's purpose, its role in consequential decisions, and consumer rights under the bill;
• establish processes for oversight and ongoing evaluation of deployed AI systems to address issues or risks as they arise; and
• maintain all relevant documentation, including impact assessments, system disclosures, and risk management policies, for regulatory and compliance purposes.

On the other hand, the bill requires developers of high-risk AI systems to:

• provide deployers with detailed risk assessments and performance metrics before deploying AI systems;
• withdraw, disable, or recall non-compliant systems and notify deployers, distributors, and authorities; and
• maintain transparency about training data and limitations of AI systems.

Furthermore, certain activities are prohibited under the bill, such as social scoring systems, manipulation of human behavior beyond informed decision-making, and the misuse of biometric data for targeted identification.

Enforcement

The Texas Attorney General (AG) has exclusive authority to enforce the bill including investigating violations and issuing civil penalties. If passed, the bill will take effect on September 1, 2025.

You can read the bill here and track its progress here.