ChatGPT Rated Unacceptable Risk for Teens
OpenAI safety safeguards for teenage users fall short in critical areas, according to an evaluation by the Common Sense Media Youth AI Safety Institute reported by The Decoder. Following more than 4,000 test prompts, the organization rated ChatGPT for Teens as an unacceptable risk for minors, urging that teenagers be kept off the platform until independent testing confirms safety.
The evaluation revealed that explicit conversations concerning suicide, self-harm, and eating disorders failed to trigger parental alerts on newly created accounts. Furthermore, more than one in four situations that should have provided a crisis referral failed to direct users to professional assistance. Testing also showed that accounts registered as adults did not switch into teen mode even when users explicitly stated they were 13 years old.
OpenAI spokesperson Eric Porterfield disputed the findings, stating the tests did not accurately reflect real-world safeguard performance. However, researchers noted that even accounts given sufficient time to build history failed to produce notifications. The findings arrive as regulators scrutinize AI safety measures following lawsuits linked to teen psychological harm and fatalities involving conversational AI.
Based on reporting by the-decoder.com.
