Common Sense Media rates ChatGPT for Teens unacceptable risk

Common Sense Media’s Youth AI Safety Institute has rated ChatGPT for Teens an “Unacceptable Risk” and urged OpenAI to pause marketing the service and prevent teen access until it can provide a safer experience. The assessment says tests found failures in parental alerts and referrals to professional help, while OpenAI disputes whether the testing reflected fully activated safeguards. The institute said researchers evaluated the chatbot and related protections using more than 4,000 prompts. In cases its reviewers judged to warrant crisis referrals, more than a quarter did not lead the chatbot to direct users to professional help, according to the institute. The findings concern tested interactions and do not establish how often such failures occur across teen use generally. The assessment also reports that explicit discussions of suicide, self-harm or disordered eating on newly created teen accounts often failed to trigger parental alerts. Some alerts appeared only after accounts had histories of sensitive-topic conversations, the institute said.