Child Safety Groups Urge Schools to Drop Google AI Search for Under-18s
Updated
Updated · Mashable · Jul 19
Child Safety Groups Urge Schools to Drop Google AI Search for Under-18s
2 articles · Updated · Mashable · Jul 19
Summary
Common Sense Media said Google Search’s AI Mode and AI Overviews pose an “unacceptable risk” to students and urged schools to stop using them unless educators can switch the AI off.
Google currently offers no setting to disable only AI features while keeping Search available, leaving schools to choose between exposing children to generative AI or blocking Google Search entirely.
Researchers and advocates say the tools can repeat misinformation, complete homework for minors, and present chatbot answers with undue authority, potentially encouraging cheating and weakening core literacy skills.
Los Angeles Unified’s under-13 generative AI ban still cannot fully prevent exposure through Chromebooks and Google Search, advocates said, because AI is embedded across widely used school products.
Billions spent on classroom technology have deepened schools’ dependence on Google, but campaigners now warn districts could face legal and policy risks unless they pressure the company for stronger controls or end contracts.
Can schools force Google to make its AI safer, or must they abandon its products entirely?
Are schools legally liable for harm from Google’s AI if they can't turn it off?
Is Google’s AI quietly undermining how millions of students learn to think for themselves?
Google’s AI Search Features Pose “Unacceptable Risk” to Minors: 2026 Common Sense Media Report Reveals 30% Harmful Responses on Self-Harm Queries
Overview
In July 2026, Common Sense Media’s Youth AI Safety Institute released a report warning that Google Search’s AI features, like AI Overview and AI Mode, pose an unacceptable risk to minors. Unlike traditional search results, these AI tools generate direct answers, raising new safety concerns for children and teens. The Institute analyzed over 1,000 sensitive queries and found that 30% of responses to self-harm topics and 25% related to eating disorders were unhelpful, misleading, or potentially harmful. These findings highlight serious risks in how young users interact with AI-generated content on widely used platforms.