OpenAI makes ChatGPT less 'human' for teens in new safety update
BBC News ● Covered by 3 sources
OpenAI is rolling out teen-safety tweaks for ChatGPT, letting under-18s mute its human-sounding voice and get break reminders. It's a nod to fears that young users might start treating the chatbot like a real, conscious friend.
Based on reporting by BBC News — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI is tightening the rules around how teenagers experience ChatGPT, starting with a feature that lets under-18 accounts turn off the chatbot's human-like voice responses entirely. The idea is to strip away some of the illusion that there's a person, or something person-like, on the other end of the conversation.
Alongside that, the company is adding regular nudges telling young users to step away from the screen. The reminders spell it out plainly: this is AI, not a friend, and whatever you're doing can wait. It's a small but pointed intervention, aimed at habits rather than content.
OpenAI says none of this was triggered by a specific case of a teenager believing ChatGPT was alive. But the timing lines up with a broader unease spreading through the industry. Mustafa Suleyman, who runs AI at Microsoft, said last year that "seemingly conscious" AI products kept him awake at night, worried about what they might do to how people relate to technology and each other.
The new teen accounts will kick in automatically for anyone who signs up and identifies as a teenager, and also for users OpenAI itself flags as likely under 18. The company hasn't explained how that detection actually works, which leaves a fairly big question mark hanging over enforcement. Under-13s remain barred from the product outright.
In its blog post, OpenAI framed the changes as recognition that teenagers need protections suited to where they are developmentally, ones that push them back toward real relationships rather than deeper into the app. That language, and the overall structure of the rollout, echoes the playbook social media companies have used for years when building out their own teen modes.
My take — AI-written commentary, not fact-checked reporting
Muting the voice and adding break reminders is a sensible start, but it's a soft patch on a much bigger problem: nobody outside OpenAI knows how it plans to tell a teenager from an adult. Social networks spent a decade proving that age-detection promises rarely hold up under real-world use, and AI companies shouldn't get a pass just because the product is newer. If OpenAI wants credit for taking this seriously, it needs to show its work, not just its intentions.”}}(note stray char)}{
Read more about this at: BBC News
