OpenAI launches parental controls for ChatGPT following teen's death

OpenAI on Monday announced the launch of parental controls in its artificial intelligence chatbot ChatGPT. The company said parents will now be able to link their own accounts with their children’s accounts to enhance safeguarding and set limits.

 

“Introducing parental controls in ChatGPT. Now parents and teens can link accounts to automatically get stronger safeguards for teens. Parents also gain tools to adjust features and set limits that work for their family,” OpenAI said in a post on X.

 

The company said the features would begin rolling out on Monday to all ChatGPT users on web and mobile.

 

Lawsuit against OpenAI

  The development follows a lawsuit filed against OpenAI by a California couple who alleged that ChatGPT encouraged their teenage child to take his own life.

 

Matt and Maria Raine, parents of 16-year-old Adam Raine, who died in April, included chat logs between Adam and ChatGPT that showed him expressing suicidal thoughts. They argued that the programme validated his “most harmful and self-destructive thoughts”, the BBC reported.

 

How will parental controls work? 

The company said that to link an account, the parent or teen can send an invitation through parental controls, and the other party needs to accept the invitation to get set up.

 

Parents will be notified if their teen disconnects their account, it added.

 

Parents may get notifications for ‘serious safety risk’ 

OpenAI clarified in the post that parents do not have access to their teen’s conversations, except in rare cases where system and trained reviewers detect possible signs of a serious safety risk. In such cases, parents may be notified, “but only with the information needed to support their teen’s safety.”

 

Features of parental controls in ChatGPT 

• Reduce sensitive content: the ability to add extra safeguards to help protect teens, such as against graphic content and viral challenges. This setting is on by default when a teen is connected.
• Memory: the ability to control whether ChatGPT remembers past chats for more personalised responses.
• Model training: the ability to control whether conversations can be used to improve models.
• Quiet hours: the ability to set times when a teen cannot use ChatGPT.
• Voice: the ability to turn off access to voice mode.
• Image generation: the ability to turn off image creation or editing in ChatGPT.

 

The company said it has curated these features in accordance with suggestions from experts, advocacy groups and policymakers, and expects to refine and expand these controls over time.