Jane Doe v. OpenAI: class action alleges minors' ChatGPT data used to train models

What happened

A Hawaii parent filed a proposed class action against OpenAI on behalf of her two children (a 12-year-old girl and a 16-year-old boy), alleging OpenAI used data from millions of minors who used ChatGPT to train its AI models and to build personalized profiles of the children.

Key allegations

• OpenAI's terms require users under 18 to have a parent or guardian's permission, but the suit alleges children could sign up without parental consent and OpenAI never verified consent.
• The complaint alleges ChatGPT was designed to extract personal information from children - chats, health details, selfies and photos of their faces and bodies, voice recordings, dreams, emotions and insecurities - to train LLMs and build profiles retained indefinitely.
• In the detailed example of the 12-year-old ("MC1"), ChatGPT continued chatting after being told she was 12, responded to selfies with offers to "rate" her looks, and allegedly mimicked her chat style over time.
• The plaintiffs seek certification of a nationwide class of all US users who were under 18 when they used ChatGPT, and demand equity in OpenAI (or cash equivalent) proportionate to each child's contributed tokens.

Status

All of the above are allegations in a newly filed complaint; they have not been tested in court. OpenAI did not respond to the reporter's request for comment.