California and Texas frontier/AI-governance laws take effect
Texas's law bans specific AI uses such as generating CSAM or discriminating against protected classes, rather than regulating frontier models by compute threshold like California's.
- Government & policy
- Notable
Two state AI laws took effect on 1 January 2026: California’s Transparency in Frontier AI Act (SB 53), signed the previous September, and Texas’s Responsible Artificial Intelligence Governance Act (RAIGA). The pair represent different regulatory approaches — California’s transparency-and-safety-protocol regime targets frontier developers above a compute and revenue threshold, while Texas’s law instead prohibits specific uses of AI regardless of model size, including generating child sexual abuse material or deceptive deepfakes, encouraging self-harm or violence, and intentionally discriminating against a protected class. Texas’s law offers developers an affirmative defence where they follow a recognised risk-management framework such as NIST’s.
The laws took effect days after President Trump signed an executive order, “Ensuring a National Policy Framework for Artificial Intelligence,” on 11 December 2025, which sought to establish a federal framework pre-empting state AI regulation the administration judged inconsistent with it and directed the Attorney General to form a litigation task force to challenge conflicting state measures. The timing set up a direct test of how far a federal order could constrain state law without new legislation from Congress.
The near-simultaneous effective dates of a Democratic-led state’s transparency mandate and a Republican-led state’s narrower, conduct-based prohibition — arriving just as the federal government moved to displace both — captured the fragmented state of US AI governance heading into 2026: broad agreement that some regulation was needed, continuing disagreement over its form, and an unresolved federal-state conflict over who gets to set the rules.