Texas Charts Independent Path on AI Regulation
The Texas Responsible Artificial Intelligence Governance Act (TRAIGA) has been passed by the Texas legislature and is currently awaiting Governor Greg Abbott’s signature. This legislation aims to regulate the development and use of artificial intelligence (AI) across both the public and private sectors, with an anticipated effective date of January 1, 2026, if signed into law.
TRAIGA represents Texas’s most thorough attempt to establish oversight over AI technologies, amidst a backdrop of national discussions regarding the implications of machine learning in daily life. This movement aligns with similar state-led initiatives in places like Colorado, Utah, and California, all seeking to impose limits on various applications of AI, even as federal lawmakers consider actions that could strip states of such regulatory authority.
Legislative Context
The passage of TRAIGA by the predominantly Republican Texas legislature marks a significant departure from the previous positioning of the Trump administration and a Republican-controlled Congress, which sought to centralize authority over AI regulation. This shift suggests a growing ideological divide between federal and state perspectives on AI governance, potentially leading to the rejection of federal legislation that aims to prohibit state-level AI regulations.
Contents of TRAIGA
Originally conceived as a broad risk-based framework influenced by the European Union’s AI Act, TRAIGA underwent substantial modifications during lengthy political negotiations and industry lobbying, resulting in a more focused scope. The current version does not outline a tiered risk model for AI systems or obligations. Instead, it emphasizes the prohibition of certain explicitly harmful uses of AI, the reinforcement of civil rights protections under existing laws, and the establishment of safeguards against biometric misuse and behavior manipulation.
Despite its narrowed focus, TRAIGA introduces significant obligations for developers, deployers, and government users of AI technologies in Texas. Notably, the bill prohibits the development or deployment of AI systems that discriminate against individuals based on protected characteristics, aligning with federal and state civil rights laws.
Real-World Implications
TRAIGA explicitly addresses behaviors that manipulate individuals’ constitutional rights, promote violence, or facilitate illegal activities through AI systems. Moreover, the bill targets the creation of AI-generated sexually explicit content and applies strict regulations on biometric data collection. For instance, biometric identifiers like fingerprints and facial recognition data cannot be collected from publicly available online media unless the subject has consented to its publication.
Exceptions exist for certain entities, such as financial institutions and companies that utilize biometric data exclusively for training AI systems without deploying them for identification purposes. However, government agencies face stricter limitations, prohibited from using AI to identify individuals based on non-consensual biometric data collection that violates constitutional rights.
Enforcement and Oversight
The enforcement of TRAIGA is designated solely to the Texas Attorney General, who holds the authority to take civil actions against violations. Individuals affected by AI misuse do not possess a private right of action under this act; instead, they can submit complaints to the Attorney General’s office for investigation. The penalties for non-compliance can reach up to $200,000 for uncorrectable violations, along with daily fines of up to $40,000 for ongoing infractions.
To facilitate regulatory compliance, TRAIGA includes a provision for an AI Sandbox, allowing businesses to test AI systems within a regulatory safe harbor for up to 36 months. Companies must provide quarterly updates on their performance and user feedback to regulators during this period, balancing innovation with public protection.
Future Considerations
TRAIGA’s viability is contingent on Governor Abbott’s decision and the ongoing federal legislative landscape. A pending federal budget reconciliation bill includes a proposed 10-year moratorium on new state AI laws, which could potentially block TRAIGA from taking effect and create tension between state sovereignty and national regulatory consistency.
In summary, TRAIGA embodies a significant and measured approach to AI regulation in Texas, addressing immediate risks while allowing for future growth and policy evolution. The legislation strikes a balance between the need for innovation in the tech sector and the imperative for public accountability and ethical standards in AI utilization.
