
Watchdog flags risky gaps in Google's AI search for kids
Common Sense Media's Youth AI Safety Institute finds Google's AI-powered search can miss crisis signals and even surface harmful content for teens, with examples where the AI failed to detect suicidal ideation or disordered eating, steered users toward unsafe forums, and encouraged behaviors like deepfakes or homework shortcuts. The researchers tested 2,600 queries from SafeSearch-enabled accounts ages 11–15 and found cases where the AI did not provide appropriate help or safety resources. Google disputed many findings, saying results aren’t reproducible and that safety guardrails exist. The report adds that a product so central to kids’ devices should be held to higher standards.









