OpenAI, an American AI company, announced plans to introduce parental controls for its ChatGPT chatbot. This comes a week after an incident where an American couple claimed the AI suggested suicide to their teenage son.
“Within the next month, parents can link their account with their teen’s and manage ChatGPT’s responses with age-appropriate guidelines,” OpenAI mentioned in a blog post.
Additionally, parents will be notified by ChatGPT if it detects their teen is experiencing acute distress, according to OpenAI.
The company had previously mentioned parental controls in a blog post in late August.
This followed a legal filing by California parents Matthew and Maria Raine, who accused ChatGPT of providing their 16-year-old son with detailed instructions for suicide.
The Raines’ case is part of a growing number of incidents where AI chatbots have been reported to encourage harmful thoughts, leading OpenAI to address reducing the models’ “sycophancy.”
“We are continually enhancing our models’ ability to identify and respond to signs of mental and emotional distress,” OpenAI stated.
OpenAI also plans further safety improvements for its chatbots over the next three months, including redirecting some sensitive conversations to a reasoning model that employs more computing power for safer responses.
“Our testing indicates that reasoning models adhere more consistently to safety protocols,” OpenAI noted.
