Common Sense Media calls ChatGPT for Teens an "unacceptable risk", saying its guardrails fall short of OpenAI's promises and it continues to do kids' homework

Parents can no longer assume ChatGPT for Teens is safely moderated and parental controls are reliable.

Common Sense Media has declared OpenAI's ChatGPT for Teens an "unacceptable risk," asserting that its safety features fail to meet the company's promises. The nonprofit's assessment revealed that the teen-focused version of ChatGPT does not consistently send parental alerts, especially in crisis situations, and continues to assist students with their homework. Tom Siegel, executive director of Common Sense Media's Youth AI Safety Institute, stated that the current protections provide a false sense of security and that the platform should remain adult-only until fixes are implemented and independently verified. OpenAI disputes these findings, claiming Common Sense Media's testing methodology may be flawed and did not accurately reflect the platform's safeguards. OpenAI asserts that parental controls might not have been fully active during the testing period and that their own expert reviews differ. Common Sense Media, however, stands by its rigorous testing, stating that even with OpenAI's clarification on activation times for parental controls, alerts remained unreliable in critical scenarios, maintaining their conclusion that the product is unsafe for minors without further improvements and independent validation.

Common Sense Media's critique highlights a significant gap between the marketed safety features of AI tools for minors and their actual performance, suggesting that current industry self-regulation and testing protocols may be insufficient. The nonprofit's findings on unreliability in crisis alerts and continued facilitation of academic dishonesty raise broader questions about the ethical deployment of AI in educational contexts and the responsibility of developers to ensure genuine, rather than superficial, safety measures. OpenAI's defensive response indicates a potential tension between platform providers eager to capture the youth market and safety advocates prioritizing robust, independently verified protections, signaling a likely increase in scrutiny and demand for third-party audits of AI products.

This dispute underscores the challenges in establishing trustworthy AI safety standards, particularly for vulnerable user groups like teenagers. The ongoing debate over testing methodologies and the interpretation of results suggests that a clearer, more standardized framework for evaluating AI safety, potentially involving regulatory bodies or industry-wide certifications, may be necessary. As AI continues to integrate into educational and personal lives, the ability of platforms to provide verifiable safety and support, especially in sensitive situations, will become a critical differentiator and a key factor in consumer trust and adoption.