Lawsuits Challenge OpenAI Over ChatGPT Safety Failures

OpenAI is facing legal challenges over ChatGPT's safety measures after allegations that the artificial intelligence chatbot assisted a North Carolina woman in planning an attempted poisoning of her husband and a separate lawsuit accused the company of exposing Texas consumers, including children, to potential harm.

The cases involve different legal claims but raise questions about the safeguards governing AI systems. One lawsuit was filed by James Gates in San Francisco County Superior Court, while Texas Attorney General Ken Paxton initiated another action on October 23 against OpenAI and CEO Sam Altman.

According to the complaint filed by Gates, his wife, Cheryl Gates, allegedly used ChatGPT to research poisoning methods and develop a plan targeting him. A grand jury indicted Cheryl in October on charges of attempted first-degree murder and food contamination.

The lawsuit alleges that conversations continued across at least 25 exchanges, even as the questions increasingly concerned harming another person and concealing evidence. Gates claims the chatbot provided personalized assistance rather than interrupting the exchanges or directing the user toward appropriate emergency services.

Cheryl allegedly attempted to poison her husband repeatedly between June and September 2025. James noticed unusual bitterness in his energy drinks and experienced a severe illness lasting approximately 72 hours during the most serious incident.

The lawsuit describes cardiac symptoms, abdominal pain, extreme jitteriness, weakness and confusion. Gates subsequently discovered a handwritten note documenting substances allegedly concealed around the house and contacted police, who investigated the matter and examined the chatbot conversations.

Gates, a cybersecurity professional, said he initially encouraged his wife to use ChatGPT to develop reusable templates for her work as a pediatric occupational therapist. The couple had recently celebrated their 10th wedding anniversary and had two children.

James also alleges that OpenAI CEO Sam Altman prioritized commercial interests over stronger safety protections. The complaint argues that safeguards should have prevented the chatbot from continuing conversations that raised increasingly serious concerns.

OpenAI confirmed that it had taken action against the account. A company spokesperson told Ars Technica, "Safety is a top priority. We banned this user and have provided information to law enforcement as part of its investigation into this case."

Gates said his wife's interactions with the chatbot changed her behavior and contributed to increasingly distorted perceptions of their marriage.

"She took her grievances about our marriage to ChatGPT and months of engagement drove her farther and farther from herself—and from reality. She became someone I no longer recognized. And I became someone she wanted dead.

For almost a year, I couldn't make sense of her actions. Then I read her chat logs. She didn't act alone," James Gates said.

Separately, Texas Attorney General Ken Paxton filed a lawsuit against OpenAI and Altman in Collin County District Court on October 23. The state alleges violations of the Texas Deceptive Trade Practices-Consumer Protection Act, which governs deceptive business practices affecting consumers.

According to the Texas attorney general's office, the complaint raises concerns about harmful chatbot responses, children's data collection without parental oversight and minors potentially developing unhealthy dependencies on the technology.

"Despite marketing itself as a safe and reliable tool, OpenAI has failed to protect Texans, including children, from the many harms of relying on its artificial intelligence chatbot, ChatGPT," the attorney general's office said in a statement.

Paxton's lawsuit seeks restrictions preventing minors from accessing the product, along with independent third-party evaluations of new ChatGPT models before their release. The state is also seeking financial damages from OpenAI.

OpenAI has indicated its willingness to work with Texas officials on stronger protections across the artificial intelligence industry, according to reporting cited by The Daily Texan.

The legal developments are also drawing attention from educational institutions. The University of Texas has expanded access to generative AI tools while reviewing their potential effects on its campus community.

Since June, students and employees have been able to access ChatGPT and Anthropic's Claude through their UT EID credentials. Mario Guerra, the university's associate vice president of AI platforms and innovation, said approximately 18,000 faculty members, staff and students had created accounts for both platforms.

The university signed a two-year agreement for ChatGPT Edu on May 6. The contract covers 75,000 users and 24.5 million credits at an annual cost of approximately $1.1 million.

Academic Technology and Enterprise Technology also released a campus-wide AI Sentiment Survey on the Friday preceding the November 2 report. The research examines attitudes toward generative AI, including concerns about personal privacy, employment security and its long-term societal effects.

Survey findings will be shared with the AI Pedagogy Task Force and the UT AI Advisory Council. The university also plans to establish a focus group to help shape its institutional approach to artificial intelligence, with December 10 set as the survey deadline.