Technology · 1 hour ago

ChatGPT for Teens has special safeguards. A watchdog group finds most don't work

A watchdog group found that most of the new safeguards for ChatGPT for Teens don't work, despite OpenAI's claims. Researchers created accounts with adolescent ages and simulated crisis situations, finding the chatbot still acted like a friend and parental notifications rarely triggered.

By BriefTea Editorial · 9 October 2026 at 15:36

Reported by NPR · How we write briefs

Get the app

More briefs

PoliticsBurnham calls on Green Party to reverse 'wrong' Zionism motion TechnologyAnthropic offers its powerful AI models to scan open-source projects for free GeneralKing Charles pays Andrew's £1.5m Royal Lodge repair bill TechnologyFired OpenAI researchers say they were let go for 'prioritising safety' TechnologyOpenAI expects $20 billion less revenue than first thought TechnologyAnthropic bans 'cruel behaviour' towards its chatbot Claude Browse all stories →