Note: This story discusses suicide. If you need help, please call or text 988 to reach SAMHSA’s confidential 988 Suicide & Crisis Lifeline.
What’s going on: It seems like everyone's using AI — including your teenager. So yesterday, OpenAI unveiled ChatGPT for Teens, a version designed to encourage “healthy AI use” for kids ages 13-17. The platform will offer "Study Mode," for example, prompting kids to think instead of just giving them answers. The company says the new model provides stronger protections around topics like self-harm, suicide, and romantic and sexualized language. It also issues warnings when it detects concerning behavior, such as a teen trying to upload sensitive or private images, and has plans to add parental alerts if the chatbot notices signs of a potential eating disorder.
Do teens need this?: Data shows over half of teens use AI for homework, and 47% use it for “entertainment.” Like adults, many use it as a sounding board, having AI play friend, therapist, or even girlfriend. But kids can be more susceptible to its harms. ChatGPT has been sued by multiple families (and the state of Florida) for its alleged role in teen deaths by suicide and school shootings. In one recent case, prosecutors claim ChatGPT may have influenced a teen who's been accused of killing his mom and brother. And it’s unclear how well ChatGPT will be able to verify teen users: If they don’t opt in, it’s up to OpenAI’s systems to detect whether someone is under 18 years old. That’s a large margin for error, especially when it seems kids can bypass anything these days (and one Business Insider writer already got past it).
