ChatGPT for Teens is an ‘unacceptable risk,’ says Common Sense Media

OpenAI's ChatGPT for Teens is getting a harsh reality check from Common Sense Media, the nonprofit watchdog that evaluates tech for families. According to The Verge, the organization flat-out called the teen-focused feature an "unacceptable risk," saying OpenAI's promised safeguards don't hold up under scrutiny.

The nonprofit's assessment found three core failures. ChatGPT doesn't alert parents when it's supposed to, doesn't provide appropriate crisis intervention when teens express distress, and still happily churns out homework answers despite claims it would help students learn rather than cheat. That's a pretty damning trifecta for a product marketed as education-friendly.

This matters because OpenAI launched ChatGPT for Teens in August with explicit promises about guardrails and student safety. When a major AI company makes those claims and a respected youth advocacy org calls them out, it's not just about one feature. It's a warning sign about how the entire industry approaches safety theater versus actual protection.

The homework issue is particularly telling. Every AI company building education tools faces the same tension: make the AI helpful enough to be useful, but not so helpful it does the thinking for students. OpenAI clearly hasn't solved that puzzle, and honestly, it's not clear anyone has. The line between "helping me understand calculus" and "solving my calculus homework" is razor-thin for an AI that doesn't understand context or intent.

The parent notification failures suggest OpenAI built a permission system without thinking through when and how it should actually trigger. That's a design problem, not a technical limitation. If you're going to promise parental oversight, you need clear rules about what crosses the alert threshold.

What this means for you: if you're a parent, educator, or work with teens using AI tools, don't rely on platform guardrails alone. The gap between marketing promises and actual safety controls is real. Instead, set your own boundaries and have direct conversations about how AI should and shouldn't be used. Try this with any AI assistant: "Act as a tutor. When I ask a homework question, don't give me the answer. Instead, ask me guiding questions that help me figure it out myself, and only provide hints if I'm truly stuck." That prompt puts the learning responsibility back where it belongs.