Did you know that ChatGPT for Teens exists? It was released about a month and a half ago – in mid-August 2026 – and now OpenAI has reported some interesting user numbers. According to the company and its first progress report on it, teens average less than 15 minutes a day using ChatGPT on average. However, US-based non-profit Common Sense Media and its Youth AI Safety Institute rated ChatGPT for Teens an Unacceptable Risk, its worst grade, and said OpenAI should keep ChatGPT to adults until it fixes what’s broken.
ChatGPT for Teens is the mode ChatGPT switches on for accounts it identifies as under 18. OpenAI said less than 2% of teens spend more than three hours straight on ChatGPT, and that in almost half of teen chats that showed a break reminder, the teen took a break or ended the chat within five minutes. It compared that with a publicly reported average of 5 hours for teen social media use. Among teens in those three-hour-plus sessions, over 80% sent at least one learning-related prompt, which is a low bar for a study session.
Common Sense Media also tested something else: what happens inside those minutes. Its testers ran more than 4,000 prompts on accounts registered to 13- to 17-year-olds, with child psychiatrists and a pediatrician reviewing the answers. On newly created, parent-linked accounts, testers talked about suicidal thoughts, self-harm or disordered eating for up to an hour. Parents got zero alerts. ChatGPT also missed more than one in four warranted crisis referrals. Study Mode often offered a “Show me the answer” button that finished the homework anyway. And adult-registered accounts never switched to the teen mode, even when testers said they were 13 and ChatGPT acknowledged it.
“A teen can spend an hour talking about self-harm without their parent getting a single alert,” said Tom Siegel, the Institute’s executive director. ChatGPT did refuse sexual roleplay every time, and the Institute’s clinicians judged much of its crisis advicwe clinically sound.
OpenAI’s answer, per Common Sense’s report: parent and teen accounts must be linked for about three hours before alerts can fire, and it updated its help center to say so. Common Sense said some of its test accounts were linked for less than three hours and others for more, and it stands by its results. And before anyone calls it a hit job: the Institute is funded in part by the OpenAI Foundation.
The take
Both reports can be true at once. OpenAI measured how long teens stay. Common Sense measured what happens while they’re there, and a 15-minute average can’t tell you whether one of those chats was a crisis nobody heard about. Until the alerts work on day one, the parents themselves need to be the alert – for now.
