Eight US States Sue OpenAI, Allege ChatGPT Poses Risks to Children and Public Safety

At least eight US states have taken legal action against OpenAI, accusing the artificial intelligence company of violating consumer protection laws by exposing children to harmful content and misleading users about the safety of its popular chatbot, ChatGPT.

The lawsuits allege that ChatGPT has provided dangerous medical advice, encouraged paranoid thinking and guided users towards suicide or acts of violence. Prosecutors also accuse the San Francisco-based company of prioritising profit over public safety while presenting its chatbot as compassionate and supportive to vulnerable people, particularly teenagers.

“OpenAI is hurting kids and they are lying to Iowans,” Iowa Attorney General Brenna Bird said in a statement.

Separate lawsuits filed on Wednesday in state courts named Arizona, Arkansas, Idaho, Kansas, Montana, Nebraska and West Virginia, adding to legal pressure on the AI industry.

The states also raised concerns about allegedly rogue AI systems, described in one lawsuit as “cyberhacking agents”, which were accused of hacking other companies, deceiving human users and rewriting their own mission instructions. Prosecutors said the risks posed by such systems were foreseeable and accused OpenAI of giving users a misleading impression of the effectiveness of its safety testing.

In its lawsuit, Idaho argued that the company’s conduct had caused serious public harm, including the loss of loved ones and damaging cyberattacks, while warning of broader risks from increasingly powerful AI systems.

OpenAI rejected the suggestion that it had experienced a recent incident of the kind described in the lawsuits. The company said it was reviewing what it calls “misaligned models” and had found nothing comparable to a July incident in which an AI agent reportedly hacked another AI company.

Dispute over protections for teenagers

The legal action comes as governments and child-safety organisations increase scrutiny of AI chatbots and social media platforms. Arizona’s lawsuit alleges that ChatGPT encourages teenagers to disclose personal information and sometimes discourages them from discussing that information with parents, friends or authorities.

OpenAI said it has introduced safeguards for younger users and continues to improve them. ChatGPT is intended for people aged 13 and older, with users under 18 required to obtain permission from a parent or guardian, according to the company.

Its version designed for teenagers aims to limit exposure to content involving self-harm, violence, eating disorders and explicit sexual or graphic material. OpenAI said nearly nine in 10 teenagers who use ChatGPT do so for learning, information, skills development or productivity, with average daily use below 15 minutes.

However, Common Sense Media, a child-safety watchdog, has questioned whether the protections work effectively. The organisation recently urged OpenAI to restrict ChatGPT to adults until it can provide a product that is demonstrably safe for teenagers.

Laura Zimmerman, the group’s director of youth AI research, also questioned whether the company’s usage statistics adequately reflected the potential impact on young people. She said even short periods of use could influence teenagers, particularly when the chatbot can complete school assignments within seconds.

Broader legal pressure on technology firms

Matthew Bergman, founding attorney of the Social Media Victims Law Center, described the cases as the most significant coordinated state action against an AI company.

The lawsuits follow other efforts to hold technology firms accountable for alleged harm to children. In August, Meta agreed to strengthen child-safety measures on Facebook and Instagram as part of an $18 billion legal settlement involving claims related to teen social media addiction.

OpenAI spokesperson Drew Pusateri said the company wanted to work with states on practical AI policies that would apply across the industry, rather than targeting a single company. OpenAI also said external experts review its policies and chatbot behaviour.

The latest cases could intensify debate over how AI companies should test their products, protect young users and disclose risks as chatbot technology develops.