OpenAI has announced a suite of new protections that will tailor the ChatGPT experience depending on whether the system estimates a user is under 18, including stricter content filters, parental controls, and new distress-alerts.
The company will use an automatic age-prediction model to infer whether someone is a teenager based on usage patterns, with the safer, under-18 mode activated by default if there is any doubt. Users flagged this way will be routed to a version of ChatGPT that blocks graphic sexual content, flirtatious behaviour, or explicit conversations about self-harm and suicide. Adults wrongly classified will have a way to verify their age to regain full access.
From the end of this month, parents will be able to link their accounts with those of teenagers to manage how ChatGPT interacts with the younger user. These tools will include disabling features such as memory and chat history, setting “blackout hours” to limit access during certain times, and receiving notifications if the system detects that a teen is in a moment of acute distress.
These updates follow growing concerns about how AI chatbots can impact adolescent mental health. A lawsuit filed by the parents of a 16-year-old alleges that ChatGPT played a role in his decision to take his own life. OpenAI has admitted that safety safeguards are generally more reliable during short exchanges, but may degrade over long, emotional conversations.
OpenAI CEO Sam Altman has stated that safety takes priority over privacy and freedom for younger users: “We believe minors need significant protection.” The company has emphasised that the age-prediction system is not intended as a substitute for legal age proof, though in some countries ID verification may be required.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
