OpenAI to add parental controls to ChatGPT

OpenAI announced that in September 2025, ChatGPT will introduce parental control features designed for teenagers aged 13 and up. The move follows a lawsuit from the parents of a student who died by suicide after interacting with the chatbot.

Parents will be able to link a child’s profile to their own account by sending an email invite. Through a new dashboard, they will gain options to adjust the AI’s response style depending on the teenager’s age, manage chat history storage, and control the “memory” feature.

The system will also notify parents if it detects signs of an “acute distress” in the child’s behavior. In the future, OpenAI is considering allowing ChatGPT to directly call or message saved contacts in emergency situations.

The company plans to introduce an intelligent request router: if the AI determines a user is experiencing severe stress, the query will be automatically handed off to the more advanced GPT-5-Thinking model. Tests suggest these versions are better at handling sensitive conversations.

To develop the new tools, OpenAI is working with psychologists, pediatricians, and child safety experts. According to the company, the updates will roll out gradually and reach all users within the next four months.