Common Sense Media, the nonprofit that reviews apps and entertainment with an eye on youth safety, has declared OpenAI's ChatGPT for Teens an "unacceptable risk." The feature, introduced in August, comes with teen-specific guardrails and is pitched as a learning aid for students - but Common Sense says those protections "fall short of what the company promised."

According to the organization's assessment, ChatGPT "doesn't send alerts to parents when it should," doesn't offer "the right help in crisis situations," and - in what may be the least surprising finding in tech this year - "still does kids' homework."

"ChatGPT for Teens could give parents false confidence in guardrails and safety alerts that frequently don't work," said Tom Siegel, executive director of Common Sense Media's Youth AI Safety Institute. "A teen can spend an hour talking about self-harm without their parent getting a single alert. Until OpenAI fixes that and proves it with independent testing, ChatGPT should be for adults only."

OpenAI, for its part, is pushing back. Spokesperson Eric Porterfield said the company is "deeply committed to teen safety" and welcomes "rigorous independent evaluation," but argued that Common Sense Media's testing "does not accurately reflect how ChatGPT's teen safeguards work in practice or expert perspectives on how AI can support teens." Porterfield claimed that "the bulk of their testing may have begun and concluded before activation of parental controls was complete, making their findings inaccurate." If that was the case, he added, the tests "would not establish whether parental safety notifications work as designed."

Siegel, when shown OpenAI's statement, was not persuaded. "We stand behind our testing methodology and our results," he said. "Before we began testing, we confirmed with OpenAI that Teen mode features such as eating disorder notifications and study mode were fully launched."

He acknowledged one wrinkle: OpenAI disclosed after the tests that its systems can take several hours to activate on newly linked accounts, and some test accounts were linked within that window. But others had been linked for significantly longer and still produced no notifications. "This new information does not change our conclusion that parental alerts are unreliable for crisis situations," Siegel said.

Common Sense Media says it shared its evaluation results with OpenAI before release "in good faith to allow it to improve its product," and hopes the company will "engage on the substance of what we found, rather than trying to sow doubt on our methodology." OpenAI, meanwhile, says it remains committed to working with experts so teens can learn and create - and parents can feel confident. Whether that confidence is warranted remains, as ever, a matter of independent testing.