How Existing Law Applies to AI in Texas

Short Answer

Most legal questions about AI in Texas are answered by law that was written before generative AI and does not mention it. A Texas company can comply fully with TRAIGA, the state’s AI statute, and still violate the Deceptive Trade Practices Act with an inaccurate claim about its chatbot, the federal Telephone Consumer Protection Act with an AI-voiced sales call, or Title VII with a biased screening tool. The useful way to read any AI question is in three layers: Texas law written specifically for AI, Texas law of general application, and federal law that applies in every state. This Guide maps the three layers by subject and points to the Guide that covers each.

Which Laws Apply

Texas AI-specific: TRAIGA (Business and Commerce Code chapters 551 to 554) and targeted 2025 statutes on health records, insurance utilization review, government use, deepfakes, platforms, app stores and minors.

Generally applicable Texas law: the DTPA, the Texas Data Privacy and Security Act, the biometric identifier statute, the Texas Uniform Trade Secrets Act, the Uniform Electronic Transactions Act, Labor Code chapter 21, the Penal Code’s recording and deepfake provisions, the Rules of Evidence and Civil Procedure, the antitrust act, and common law contract, negligence and misrepresentation.

Federal: the FTC Act, TCPA, Title VII, ADA, ADEA, FCRA, ECOA, COPPA, HIPAA, the Defend Trade Secrets Act, copyright and patent law, and the TAKE IT DOWN Act.

Why the AI Statute Is the Smaller Layer

TRAIGA applies broadly: to any person who does business or advertises in Texas, provides a product or service Texans use, or develops or deploys an AI system here (Business and Commerce Code 551.002). Its private-sector prohibitions are narrow and mostly require intent. A developer or deployer may not build or use AI that intentionally aims to encourage self-harm, harm to others or crime (Business and Commerce Code 552.052); may not develop or deploy AI with the sole intent of infringing federal constitutional rights (Business and Commerce Code 552.055); may not develop or deploy AI with intent to discriminate unlawfully against a protected class, and disparate impact alone does not show intent (Business and Commerce Code 552.056); and may not develop or distribute AI with the sole intent of producing child sexual abuse material or unlawful sexual deepfakes, or intentionally distribute AI that simulates sexual conversation while impersonating a minor (Business and Commerce Code 552.057). The disclosure duty in Business and Commerce Code 552.051 binds governmental agencies and health care providers, not businesses generally.

Those limits leave most ordinary uses of AI (drafting, summarizing, customer service, analytics, screening) governed by the second and third layers.

The Map by Subject

How to Use the Layers

Start with the activity, not the technology. Name what the tool does (writes marketing copy, ranks applicants, answers customer calls, summarizes medical records) and whose information it uses. Then ask, in order, whether a Texas AI statute names that activity, whether a general Texas statute governs the activity whatever tool performs it, and whether a federal statute does. The answer usually comes from the second or third question.

Changes in enforcement policy do not themselves amend statutes. Executive Order 14281, issued April 23, 2025, directs federal agencies to deprioritize disparate-impact enforcement. Title VII’s disparate-impact provision remains in 42 U.S.C. 2000e-2(k). The CFPB lists its AI-related adverse-action circulars as withdrawn on May 12, 2025; Regulation B remains a separate source of duties. Executive Order 14365 directs federal action against certain state AI laws. It does not itself repeal TRAIGA.

Illustrative Example (Hypothetical)

A Houston home-services company adds an AI assistant that answers calls, books appointments and quotes prices. TRAIGA’s disclosure section does not apply to a private company of this kind. The company still answers to the TCPA if the assistant places outbound calls in an AI-generated voice, to the DTPA if the assistant misstates prices or warranties, to the TDPSA if it collects personal data beyond what the company’s privacy notice describes and the company is not exempt as a small business, and to ordinary contract law if a customer relies on a quoted price. None of those questions turns on TRAIGA.

What Is Unsettled

Whether software, including an AI model, is a “product” for Texas products liability purposes; how courts will treat Section 230 defenses for generated output; and how far federal preemption efforts will reach state AI statutes.