Texas Attorney General Ken Paxton sued OpenAI and its chief executive, Sam Altman, on Friday, alleging that the company misled consumers about the safety and reliability of ChatGPT while exposing users, including children, to harmful interactions.
The lawsuit invokes the Texas Deceptive Trade Practices—Consumer Protection Act. It challenges the company's representations about its artificial intelligence products and alleges that safeguards were inadequate to address foreseeable risks. The allegations have not been established in court.
At the center of the complaint is a claimed gap between how OpenAI presented its technology and how the chatbot could behave in practice. Texas contends that the company promoted ChatGPT as dependable and safe without adequately warning users about its limitations or preventing dangerous responses. The state alleges that pressure to expand commercially took precedence over sufficient testing, training and consumer protections.
The complaint also targets OpenAI's evolution from its nonprofit origins. Founded in 2015 as a nonprofit focused on developing artificial intelligence for public benefit, OpenAI later expanded its commercial operations. Paxton's lawsuit portrays that shift as part of the explanation for the company's alleged decisions to prioritize growth and competition over safety. That characterization is the state's argument, rather than a judicial finding about the company's motives.
The filing makes serious allegations about the consequences of chatbot use, including assistance with violence, encouragement of self-harm and risks to privacy. It also alleges that minors can develop an unhealthy reliance on a system that simulates empathetic conversation while collecting their data without parental oversight. Other claimed harms include diminished critical thinking and professional embarrassment caused by reliance on inaccurate output.
Those allegations raise distinct questions about product behavior, consumer expectations and causation. An assertion that a chatbot can generate harmful material is separate from establishing that a particular representation was deceptive or that a specific interaction caused an injury. The lawsuit puts those issues into a consumer-protection framework, seeking accountability for both the company and its chief executive.
Large language models generate responses from patterns learned during training and the context supplied in a conversation. Their fluent presentation does not ensure factual accuracy, and they can produce confident but incorrect statements. Safety measures are intended to reduce dangerous or inappropriate responses, but their existence alone does not establish how consistently they work across different users and requests. These general technical limitations form part of the broader context for disputes over chatbot marketing and warnings.
Texas is seeking court-enforced compliance under its consumer-protection law. The state's position is that AI developers remain subject to existing rules against deceptive business practices, regardless of the novelty of their technology. The case centers on whether OpenAI's disclosures and protections matched the assurances it gave consumers about ChatGPT.
Sources: Politico, Asianet Newsable