OpenAI adds parental controls to ChatGPT after teen's death

The controls will let parents and teens link accounts, decide whether ChatGPT remembers past chats, and limit teens' exposure to sensitive content, the company said Monday.

OpenAI said it will also try to notify parents when it detects “signs of acute distress” in a child. "If our systems detect potential harm, a small team of specially trained people reviews the situation," the company said.

The chatbot maker added that its system isn't perfect and "might sometimes raise an alarm when there isn’t real danger."

The rollout comes after allegations by a California family that ChatGPT played a role in their son's death. The lawsuit, filed in August by Matt and Maria Raine, accuses OpenAI and its chief executive, Sam Altman, of negligence and wrongful death. They allege that the version of ChatGPT at that time, known as 4o, was “rushed to market … despite clear safety issues”.

Their son, Adam, died in April, after what their lawyer, Jay Edelson, called “months of encouragement from ChatGPT.” Court filings revealed conversations he allegedly had with the chatbot where he disclosed suicidal thoughts. The family allege Adam received responses that reinforced his “most harmful and self-destructive” ideas.

[Editor's note: The national suicide and crisis lifeline is available by calling or texting 988, or visiting 988lifeline.org.]

ChatGPT’s Monday blog post said parents will be allowed to set quiet hours so as to block access at certain times of day. They will also be able to disable voice mode, and stop the app from generating images. Parents will not be able to access their children’s chat transcripts.

"We’ve worked closely with experts, advocacy groups, and policymakers to help inform our approach — we expect to refine and expand on these controls over time," the company said in a statement.

Robbie Torney, senior director of AI Programs at nonprofit Common Sense Media, said the controls are “a good starting point,” but added they are “”just one piece of the puzzle” on online safety. 

“They work best when combined with ongoing conversations about responsible AI use, clear family rules about technology, and active involvement in understanding what their teen is doing online,” he said.

Two weeks ago, Altman said the company will at some point try to detect underage ChatGPT users. “If there is doubt, we’ll play it safe and default to the under-18 experience,” Altman said. “In some cases or countries we may also ask for an ID; we know this is a privacy compromise for adults but believe it is a worthy tradeoff.”

Researchers have documented how easy it is to circumvent limits set by chatbot companies. Age-verification rules are also known to be easily bypassed.

—Niamh Rowe and Hannah Parker contributed to this article.