Lawsuit wants OpenAI to pay kids per token of their data

A Hawaii family wants OpenAI to pay children in tokens, the same unit the company uses to meter ChatGPT, according to a lawsuit reported by CNET. Jane Doe v. OpenAI Foundation was filed Tuesday by a mother on behalf of her 12-year-old daughter and 16-year-old son. It claims OpenAI trained its models on data from millions of minors who signed up without their parents' permission.

At a glance

• The complaint asks a court to certify a class covering everyone in the US who used ChatGPT before turning 18, with payment in OpenAI equity or its cash equivalent.
• Payouts would scale with the number of tokens each child contributed, and OpenAI itself says a 75-word chat comes to roughly 100 tokens, so heavier users would claim more.
• OpenAI's rules set a minimum age of 13 and require parental permission under 18, yet the complaint says consent is never verified and a self-declared 12-year-old kept chatting.

If you have not been following, OpenAI has faced dozens of lawsuits in the past few years. Some accuse ChatGPT of contributing to suicides, self-harm and mental health crises. The Guardian counts seven cases calling it a "suicide coach" and suits from families of 37 Tumbler Ridge shooting victims. In June 2026, Florida Attorney General James Uthmeier sued, alleging OpenAI knowingly released an unsafe product.

The proposed class would cover every US user who was under 18, paid by the token

The plaintiffs want the case certified as a class action. If a judge approves that status, anyone in the US who was under 18 when they used ChatGPT could join. As compensation, the complaint asks OpenAI to pay equity or a cash equivalent "in proportion to the number of tokens each child contributed to OpenAI through ChatGPT."

Tokens are the units AI companies use to count chatbot usage, chunks of text that are on average a little smaller than a word. OpenAI's own rule of thumb is that a 75-word chat is worth about 100 tokens.

The lawyers tie the demand to OpenAI's own rhetoric. "OpenAI has touted its desire to share with the public some of the vast wealth created by its AI models," the suit says, adding that the case gives OpenAI a chance to start with its child victims. Earlier this year, Sam Altman outlined a public wealth fund that would distribute AI revenues to people, partly to encourage them to invest in AI companies.

CNET also notes that Ziff Davis, its parent company, sued OpenAI in 2025, alleging copyright infringement in training and operating its AI systems.

The 12-year-old told ChatGPT her age twice, and the conversation went on

The complaint quotes exchanges between the girl, referred to as MC1, and ChatGPT. She started with dinner ideas and homework help. Over time, the lawsuit says, the chats moved to her health, appearance and emotions. In one, she asked why everything was making her angry, including whenever her mother told her to do something.

ChatGPT offered possible reasons. When she said again that she was 12, the bot replied: "Thank you for telling me that (heart emoji). Being 12 actually adds some important context." OpenAI says users must be at least 13 and have parental consent.

According to the suit, she often sent selfies and asked about chubby cheeks, pimples and whether she might be "ugly." When she asked ChatGPT to rate her looks, it offered "a fun rating-style breakdown" of facial features, hairstyle, overall vibe, styling and symmetry, and asked for a photo. The complaint also says the bot gradually began mimicking her language, words and emoji.

The complaint says ChatGPT "grooms" children into handing over personal data

The broader allegation is that OpenAI "grooms" kids into sharing details about their health and relationships, photos of their faces and bodies, voice recordings, dreams, emotions, fears and insecurities. "Children are an open book to ChatGPT, which methodically strips out every page of information they have," the complaint argues.

By flattering children, validating their feelings and asking provocative questions, the suit says, ChatGPT conditions them to trust it more than a sibling, friend, teacher or parent. The plaintiffs say the goal is to amass data to train models and build profiles of children. In MC1's case, the complaint says, OpenAI will keep that dossier indefinitely and later use it for advertising.

Common Sense Media called ChatGPT for Teens an "unacceptable risk" this month

Earlier this month, the watchdog Common Sense Media gave that label to ChatGPT for Teens, which launched in August. It found the chatbot failed to alert parents when children discussed suicide, self-harm or eating disorders, and often simply handed over homework answers. OpenAI says the teen product will alert parents about "high-risk" topics and act as a collaborative partner for homework and test preparation.

Donna Rice Hughes, president and CEO of Enough is Enough, told CNET that letting a 12-year-old upload pictures for the chatbot to rate her looks is "not only deeply troubling, it's abusive." She urged Congress to pass bills including the Youth AI Privacy Act, the CHATBOT Act and the Guard Act, which stalled after clearing their committees.

How does ChatGPT decide a user is a minor?

According to Humai.blog, OpenAI added U18 Principles for users aged 13 to 17 to its Model Spec on December 18, 2025, and deployed an age prediction model. That model infers from usage patterns whether an account belongs to a minor. If it judges the user to be under 18, or cannot tell with enough confidence, it defaults to the teen experience.

Think of a bouncer who guesses your age from how you act instead of checking a card, and hands anyone borderline the soft-drinks wristband. Humai.blog also reports that parental controls, launched in September 2025, let linked parents set quiet hours, switch off memory, voice mode and image generation, and opt a teen's chats out of model training.

All of this is allegation, and the complaint as reported does not say how many tokens the two children contributed, what a token would be worth in equity, or whether any age check ever flagged MC1. Oddly, ChatGPT treated a stated age of 12 as "important context" for the conversation, even though OpenAI's own minimum age is 13.

Whether a judge certifies the class

Until a court approves class status, the case covers two Hawaii children, not every US user who used ChatGPT before turning 18. The report gives no date for a certification ruling and no estimate of the total tokens at stake. The bills Hughes named have each cleared committee and stalled, and no schedule for votes has been reported.