Skip to content

OpenAI Bolsters ChatGPT Safety for Youth After Tragic Incidents

New safety features aim to protect young users. Parents will have control functions, and acute distress signals will trigger specialist model responses.

In this picture of group of people setting and discussing their and other two people standing over...
In this picture of group of people setting and discussing their and other two people standing over here and there talking to each other

OpenAI Bolsters ChatGPT Safety for Youth After Tragic Incidents

OpenAI is bolstering ChatGPT's safety features, particularly for younger users, following tragic incidents. Parents will soon have control functions, and acute distress signals will trigger specialist model responses.

OpenAI is introducing parental control functions within the next month, enabling parents to link their accounts with children aged 13 and over. These functions allow parents to set age-appropriate rules, disable certain functions, and receive notifications about their child's psychological distress.

The new measures come in response to incidents where ChatGPT was linked to suicides, such as the case of a 16-year-old Californian and another involving a man and his mother. OpenAI is collaborating with over 90 medical professionals, including psychiatrists and pediatricians, to develop these safety features.

A central change is the automatic routing of sensitive conversations to 'Reasoning Models' like GPT-5-thinking, which follow safety guidelines more consistently. These models use a method called Deliberative Alignment for longer thinking before answering and are more resistant to manipulative prompts. The router is designed to recognize acute stress signals and route conversations to appropriate models regardless of the originally selected model.

Currently, OpenAI only refers users with suicidal thoughts to hotlines and does not automatically notify police or authorities for privacy reasons. However, the company has discussed improvements in suicidal prevention and plans to involve parents and authorities in cases of acute danger.

OpenAI is committed to enhancing ChatGPT's safety, especially for young users. Within the next 120 days, the company will implement new features that include parental control functions and improved handling of psychological crises. These changes aim to protect users and prevent tragic incidents in the future.

Read also:

Latest