OpenAI announced on Tuesday a new mode called "ChatGPT for Teens," which will automatically limit certain conversations to better protect younger users, as scrutiny grows over the potential harms of artificial intelligence chatbots, according to The New York Times. The company said in a blog post that the chatbot will restrict high-risk conversations involving self-harm, violence, eating disorders and explicit sexual or graphic content. In some cases, ChatGPT will alert adults who have opted into parental controls and notifications, and the mode will also strengthen safeguards that prevent the chatbot from suggesting it has personal feelings.
The teen mode also includes new study tools, the Times reported. If ChatGPT detects that a teenager is working on homework, it is designed to respond with further questions rather than providing direct answers.
Detecting under-18 users
OpenAI requires users to state their age when signing up, with a minimum age of 13 under its terms of use, though the company acknowledged that some users misrepresent their age. According to the Times, OpenAI tracks more than 2,000 signals to detect whether a user is under 18, automatically activating teen mode where that is the case. A company spokesman described these signals, which can include typical login times and how long an account has existed, as "minimally privacy invasive."
Lawsuits and mounting pressure
OpenAI, which requires parental consent for teenagers to use its chatbot, has faced growing pressure over the safety, mental health and developmental effects of the platform on younger users. A number of families have sued the company, alleging their children were exposed to harmful content and subsequently engaged in self-harm or died by suicide as a result, according to the Times. In June, Florida became the first US state to sue OpenAI, accusing the company of misrepresenting the safety of its platform for minors.
Separately, parents' groups in New York City, Los Angeles and elsewhere are pressing schools to pause the rollout of AI tools for students, citing concerns that chatbots could hinder learning, spread misinformation or facilitate cheating, with some education researchers warning the tools may impede young people's critical thinking. OpenAI described its new study mode as an additional educational tool intended to support teenagers "outside the classroom when help is not always available," while keeping them engaged in an active learning process, according to the company's blog post cited by the Times.
The Times noted that OpenAI's approach mirrors a pattern long used by social media companies, introducing youth-specific modes with added safeguards while continuing to encourage use of the platform overall, an approach that relies heavily on parental controls that child-safety advocates and lawmakers have criticised as shifting responsibility for harm away from the companies themselves. Verifying users' ages has remained a persistent challenge for technology companies generally, given how easily such checks can be circumvented, the Times reported.
Source: The New York Times
