TRAIGA Applies to Physical as Well as Virtual Environments
Published
Statute in effect since January 1, 2026. Law checked through October 18, 2026.
The ancient historians report that at Syracuse, Archimedes built a claw that lifted Roman ships from the water. Whatever the engineering, the lesson was that intelligence becomes most consequential when it moves something heavy. The Texas Legislature wrote that lesson into its AI statute.
The Texas Responsible Artificial Intelligence Governance Act, House Bill 149, took effect January 1, 2026, as Chapters 551 through 554 of the Business and Commerce Code. Section 551.001(1) defines an artificial intelligence system as a machine-based system that infers from its inputs how to generate outputs that "can influence physical or virtual environments." A robot that perceives a warehouse aisle and decides where to move fits that definition as squarely as any chatbot.
Inclusion in the definition does not make TRAIGA a robot-safety code. The Act imposes targeted prohibitions and disclosure duties, places enforcement with the Attorney General, subject to a 60-day opportunity to cure, creates a regulatory sandbox, and establishes a state AI council. HB 149 also amended the biometric-identifier statute, § 503.001, which matters to any robot whose cameras can identify a face. Whether a particular deployment triggers a particular provision depends on that provision's text and on what the system is designed and used to do.
For Texas deployers, the practical step is to fold physical systems into the same AI inventory and review process used for software: identify which robots make inferences, what data they collect, and whether any use approaches a prohibited purpose or a biometric rule.
Limits
TRAIGA does not license robots, set performance standards, or displace product-liability, workplace-safety, or vehicle law.
Sources
- Acts 2025, 89th Leg., R.S., ch. 1174 (H.B. 149)
- Tex. Bus. & Com. Code § 503.001
- Tex. Bus. & Com. Code chs. 551-554