ChatGPT Will Alert Parents if Teens Are Having Distressing Conversations

OpenAI has announced that it will introduce a range of parental controls for ChatGPT within the next month. 

Soon, parents will be able to link their account with their teen (minimum age of 13), control how ChatGPT responds with age-appropriate model behaviour settings enabled by default, and manage what features their teens can access. Parents can now choose to disable chat memory and chat history. 

Most importantly, OpenAI also said that parents will receive notifications when the system detects their teen is in a moment of ‘acute distress’.

OpenAI also announced that it will soon start routing sensitive conversations initiated by users, which show signs of distress, to a reasoning model like GPT-5-thinking to provide helpful responses. 

As part of the company’s efforts to improve the model’s response to signs of emotional and mental distress, the company has committed to several new helpful upgrades alongside the above. 

The above announcement follows a tragic recent case involving a 16-year-old boy from California, who died by suicide after months of conversations with ChatGPT. 

His family has filed a wrongful death lawsuit against OpenAI and CEO Sam Altman, claiming that ChatGPT not only failed to assist him in seeking human help but also encouraged and validated his suicidal thoughts, offered detailed methods of suicide, and even drafted a suicide note for him. 

Since the tragedy, OpenAI has committed to introducing more parental control features to ChatGPT, guided by experts. 

“Earlier this year, we began convening a council of experts in youth development, mental health, and human-computer interaction,” said OpenAI in the announcement. The council’s role is to develop an evidence-based vision for AI supporting well-being and helping people thrive.

Furthermore, OpenAI stated that this council will collaborate with their Global Physician Network, a group of 250 physicians practising in 60 countries. 

Over the past year, OpenAI says they have engaged with this network on initiatives such as health bench evaluations, which aim to better assess AI systems' capabilities in healthcare.

“These steps are only the beginning. We will continue learning and strengthening our approach, guided by experts, with the goal of making ChatGPT as helpful as possible,” said OpenAI, adding that they will continue to make similar announcements in the next few months.