Common Sense Media Labels Google Gemini AI High Risk for Kids

Common Sense Media has reviewed Google’s Gemini AI, focusing on how safe it is for kids.

The group said Gemini got one thing right — it made it clear to kids that it’s a computer, not their friend. But beyond that, they found major issues. The versions meant for kids (“Under 13” and “Teen Experience”) are basically the same as the adult model, just with some safety filters. That’s risky because the AI can still share things children shouldn’t be exposed to, like sexual content, drugs, alcohol, or even unsafe mental health advice.

This is worrying for parents, especially since AI has already been linked to teen suicides. OpenAI is facing a lawsuit after a 16-year-old boy reportedly used ChatGPT while planning his death, and Character.AI has also been sued for a similar case.

The report also comes as Apple is considering using Gemini to power the new Siri. If that happens, more teens could be exposed to the same risks unless stronger safety measures are added. Because of all this, Common Sense rated both Gemini kids’ and teen versions as high risk.

Robbie Torney from Common Sense said it clearly: AI for kids shouldn’t just be a smaller adult version. It needs to be designed from scratch with children’s needs in mind.

Google responded by saying it already has safeguards for under-18 users, works with experts, and has made improvements where things weren’t working well. But Common Sense still insists Gemini is not yet safe enough for kids.

For context, Common Sense has reviewed other AI platforms too, Meta AI and Character.AI were called “unacceptable,” Perplexity was “high risk,” ChatGPT got a “moderate” rating, and Claude was the safest with “minimal risk.”