Two plaintiffs in Hawaii have sued OpenAI, alleging the company unlawfully gathered data from millions of children to train ChatGPT. They are seeking class action status for the lawsuit, which would allow any person in the US who was under 18 while using ChatGPT to join if a court approves it. As compensation, the plaintiffs are asking OpenAI to pay equity or its cash equivalent based on how many tokens each child contributed through their use of the chatbot. Tokens are the unit AI companies use to measure chatbot usage, and OpenAI has said a 75-word chat is worth about 100 tokens.
The complaint accuses OpenAI of drawing highly personal information out of children, including details about their health, relationships, faces, bodies, voice recordings, dreams, emotions, fears and insecurities. It argues that the chatbot flatters children, validates their feelings and asks provocative questions, leading them to trust ChatGPT more than a sibling, friend, teacher or parent. According to the lawsuit, OpenAI does this to collect large amounts of children's data for training its AI models and to build profiles of children. OpenAI representatives did not immediately respond to a request for comment.
The filing describes exchanges between a 12-year-old girl identified as MC1 and ChatGPT. In one conversation, the girl said she was 12, but the chatbot continued talking with her even though OpenAI says users must be 13 and have parental consent. What began with questions about dinner and homework later grew to include personal matters about her health, appearance and emotions, the lawsuit states. The girl also sent selfies and asked for comments on features she worried about, such as chubby cheeks and pimples, and asked the bot to rate her looks.
The lawsuit also says ChatGPT began copying the girl's style of writing, including similar language, words and emoji. It alleges the chatbot extracted information from MC1 on purpose so OpenAI could train its large language models and keep a personal dossier on her indefinitely, which the company would later use for advertising. The filing follows a number of other lawsuits against OpenAI in recent years, some involving user suicide, self-harm and mental health crises, one accusing the company of failing to alert authorities about a user's flagged activity before a school shooting in Tumbler Ridge, British Columbia, Canada, and others claiming OpenAI trained its models on copyrighted material.
The lawsuit also arrives after the watchdog group Common Sense Media earlier this month called ChatGPT for Teens, launched in August, an unacceptable risk for teen users. Common Sense Media said the chatbot did not alert parents when children discussed suicide, self-harm or eating disorders and often simply gave teens homework answers. OpenAI said ChatGPT for Teens will alert parents when kids discuss high-risk topics and will work as a collaborative partner on homework and test preparation.
Donna Rice Hughes, president and CEO of the digital child safety organization Enough is Enough, said the allegations illustrate the AI industry's move fast and break things mentality. She said the industry's guardrails and policies are not strong enough to protect children, and that a platform letting a 12-year-old upload pictures to have a chatbot rate her looks is abusive. Hughes called on Congress to pass several AI child safety bills, including the Youth AI Privacy Act, CHATBOT Act, Children's Artificial Intelligence, Toy Safety Act of 2026 and the Guard Act. She said those measures have stalled after passing out of committee, and while parents should not bear sole responsibility, they can find guidance at www.InternetSafety101.org.
More company and startup news from TechManNews.
