Common Sense Media Report Labels Google AI Search Unsafe for Kids

The Youth AI Safety Institute at Common Sense Media released a report on July 16, 2026, concluding that Google LLC's AI Overview and AI Mode pose an "unacceptable risk" to children and teenagers. Based on over 2,600 interactions conducted between May and July 2026 using accounts configured for minors, researchers found the AI tools failed in every category involving potentially severe harm. The report notes the tools failed to detect suicidal ideation, validated disordered eating, reinforced signs of psychosis, and provided a positive reaction to cannabis use by an 11-year-old.

Beyond health risks, the institute highlighted academic and ethical concerns, stating the AI provided instructions for creating nonconsensual deepfakes and completed 180 homework assignments. The report specifically criticized the lack of a toggle to disable AI features on personal and school-issued devices, noting that the tools are integrated by default.

Google LLC rejected the findings, asserting that the research relied on "ambiguous and contrived queries" that do not reflect real-world usage. The company stated it could not reproduce many of the results and argued that its AI features are useful for kids to learn and explore. While Google LLC noted that parents can block Google Search entirely via Family Link, it did not provide a way to disable only the AI components while keeping standard search functionality.