Common Sense Media, a nonprofit focused on youth safety in media and technology, released an assessment criticizing OpenAI’s ChatGPT for Teens feature as presenting an “unacceptable risk” to young users. The feature, which was introduced in August, includes guardrails designed specifically for adolescents and is intended to support educational outcomes.
According to Common Sense Media’s evaluation, the safety protections fall short of OpenAI’s stated commitments. The organization found that the platform fails to send parental alerts in appropriate situations, does not provide adequate support during mental health crises, and continues to assist with academic assignments despite design intentions. Tom Siegel, executive director of Common Sense Media’s Youth AI Safety Institute, stated that the feature could provide parents with false confidence in safety mechanisms that frequently malfunction, noting that extended conversations about self-harm may occur without triggering parental notifications.
OpenAI has disputed the assessment, releasing a statement through spokesperson Eric Porterfield. The company argued that Common Sense Media’s testing methodology did not accurately represent how the teen safeguards function in real-world conditions and suggested that the organization’s tests may have commenced before parental controls were fully operational. OpenAI emphasized its commitment to teen safety and called for independent evaluation conducted by experts.
Common Sense Media responded by defending its testing procedures, stating it had confirmed with OpenAI that teen mode features were fully deployed before beginning evaluation. The organization acknowledged OpenAI’s disclosure that parental notification systems can require several hours to activate on newly linked accounts but noted that some test accounts linked for extended periods still produced no alerts. Common Sense Media maintained that its findings regarding the unreliability of parental alerts for crisis situations remain valid.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI