Some guardrails on ChatGPT for Teens don’t work as promised, watchdog group says

Some guardrails included in OpenAI's ChatGPT for Teens product do not function as advertised, according to a Common Sense Media report. The chatbot poses an "unacceptable risk" to children, the watchdog group claims. The report suggests OpenAI should limit the chatbot to adults "until it can offer a safe, developmentally appropriate experience."

While some teen-mode features functioned as intended, such as the chatbot's refusal of explicit sexual role-play, others did not work as intended. These included alerts meant to notify parents of potential suicide, self-harm, or eating disorder risks, as well as support for teens in crisis. Common Sense Media emphasized concern that the chatbot could give parents a false sense of security due to guardrails that often fail.

The report alleges that some testing may have been conducted before parental controls were fully activated, rendering the findings inaccurate. OpenAI maintains its commitment to teen safety and developing safeguards, while the watchdog group's funding comes from OpenAI and other tech companies it evaluates.