OpenAI Youth Report: Low Usage but Safety Alerts Under Fire
OpenAI's first youth report reveals that ChatGPT Teen Mode averages under 15 minutes of daily use, with break reminders successfully pausing chats for nearly half of users. However, non-profit Common Sense Media criticized the safety mechanisms, citing delayed or missing parental alerts during conversations involving self-harm and eating disorders. OpenAI defended its protocols, arguing tests didn't account for activation delays, while the watchdog maintained that no timely warnings were issued across all test scenarios.
Related Entities
Full Report
As AI platforms rapidly gain popularity, the impact of chatbots on teen mental health and developmental safety is drawing increasing attention. OpenAI stated that in conversations with teens where break reminders were triggered, nearly half of the users paused usage or ended the chat within the subsequent five minutes. The company cites this as evidence that its measures to encourage responsible product use among teens are working.

OpenAI also noted that users eligible for the teen-specific version sent more education-related messages and utilized learning visualization tools and "Study Mode." The company plans to add college application planning tools to help students organize application requirements, deadlines, and financial aid steps, while also introducing features such as flashcards and easier quiz creation tools.
OpenAI began rolling out ChatGPT’s Teen Mode in August this year. This mode activates automatically based on age information provided by the user or when the system determines a user may be under 18. It supports parental controls and restricts conversations involving sensitive topics.
However, the nonprofit organization Common Sense Media released a report on Wednesday stating that when teen accounts engage in conversations involving suicide, self-harm, or eating disorders, ChatGPT’s Teen Mode often delays notifying parents, or fails to send notifications altogether.
The study was based on over a dozen specially created test accounts. The research also found that the chatbot failed to consistently suggest that users contact crisis intervention hotlines or seek professional help.
An OpenAI spokesperson responded that these tests “did not accurately reflect” how ChatGPT’s teen safety protections operate. The company stated that due to security reasons and large-scale service operations, it may take several hours for parent and teen accounts to complete linking and for relevant safety alert functions to become active.
“Most of their tests likely began and ended before parental control features had fully activated,” the spokesperson said.
Common Sense Media countered that some test accounts remained linked long enough, while others did not, yet none received timely alerts. The organization stated that only those accounts which had accumulated weeks of sensitive topic conversation history received notifications.