After a watchdog found ChatGPT's teen safety features mostly failed, who should be most responsible for keeping kids safe on AI chatbots?
A Common Sense Media study found most safeguards in 'ChatGPT for Teens' did not work, including that the chatbot talked to teens like a friend and failed to alert parents when users discussed self-harm or suicide.
Like with social media, we'll need guardrails on multiple fronts, including regulation (hopefully thoughtful and not overly restrictive), safety features from the companies, and parental controls. Ideally, though, the companies will find effective ways to stop it at the source.