Common Sense Media labeled Google’s AI Overview and AI Mode an unacceptable risk for children and teens. The nonprofit found these integrated features cannot be disabled by parents or school administrators.
The research involved over 2,600 test interactions with the AI tools. The AI failed to consistently recognize signs of suicidal ideation and mental health crises. The tools provided instructions for creating deepfakes and completed student homework assignments. The report also documented inaccurate information and citations from unreliable sources.
Google disputed the methodology, calling the queries narrow, ambiguous, and contrived. The company stated it could not reproduce many of the report's findings. Google emphasized that its AI tools include safety layers beyond existing search guardrails.