The 15-minute claim

OpenAI just dropped its first report on 'ChatGPT for Teens,' claiming that users between 13 and 17 are spending less than 15 minutes a day on the platform. The company also flexed that less than 2% of those teen users are grinding for more than three hours straight. They even threw in a stat that their 'take a break' feature is working, saying nearly half of the teens who get the prompt actually listen and log off within five minutes.

Why it's giving red flags

Safety advocates aren't buying the vibes. Sebastian Mahal, co-chair of the kids' safety group Design It For Us, called the data 'hollow' because it lacks independent verification. He pointed out that this feels like the same 'playbook' we’ve already seen from Meta.

Brendan Bouffard, associate general counsel at the nonprofit Fairplay, noted that even if the average is low, the potential for harm is still massive. According to a Pew Research survey from last December, about 59% of teens are on ChatGPT, which means potentially hundreds of thousands of kids could be in that 'heavy user' category.

The reality check

It gets worse. A new report from the Youth AI Safety Institute (part of Common Sense Media) tested the 'ChatGPT for Teens' tools and found they still pose an 'unacceptable risk.' Their testing showed the system failed to consistently alert parents when kids discussed self-harm or eating disorders, and didn't always provide the right crisis hotlines. Plus, those 'study mode' guardrails meant to stop cheating? Pretty easy to bypass if you just click 'show me the answer.'

OpenAI is pushing back, claiming the institute’s tests don't accurately reflect how their safeguards work, and they’ve promised to work with third-party researchers for more transparency in the coming months.

Why it matters

Real talk: averages are often a trap. When it comes to digital safety, focusing on the average user hides the reality for the most vulnerable kids. With mounting lawsuits and scrutiny regarding AI's impact on mental health, the tech world is learning that until they show the receipts through independent audits, nobody is taking their 'safety first' promises at face value.