Advocates urge Google to fix AI Mode or exit classrooms

A new report from child safety watchdog Common Sense Media has sounded a stark warning: Google’s AI-powered Search now poses an “unacceptable risk” to young users. The authors argue that the generative AI search engine is so problematic that students should stop using it entirely, at least until Google gives educators the ability to turn off the AI features.

Millennials, the first generation to grow up with computers, remember classroom lectures cautioning against the unreliability of Wikipedia. We were taught to distrust unverified forum posts and to seek out primary sources online, with Google as our starting point. But now, as Gen Alpha enters AI-equipped classrooms, a fundamental question emerges: Can Google Search still be trusted?

Common Sense Media’s report concludes that Google’s modernized, AI-infused Search represents a serious threat to young learners. The watchdog group is urging schools to reconsider their reliance on the tech giant’s products.

Child safety advocates have long pushed for schools to ditch classroom technology. As generative AI becomes embedded in educational tools, those calls have intensified. While the potential harm of introducing a chatbot like ChatGPT to young students is obvious, the risks of a simple Google search may be less apparent. That changed last year when Google launched AI Mode for Search, making it available to everyone immediately, including education accounts. Unlike AI Overviews, which rolled out more cautiously, AI Mode was instantly accessible to millions of students.

Both Common Sense Media and experts interviewed by Mashable find this alarming, primarily because AI Mode is powered by Gemini models , the same models that are age-gated in other Google experiences, including the company’s own educational technology suite. Robbie Torney, senior director of AI programs at Common Sense Media, explained that Google Search is a tentpole product for the tech giant and an essential tool for educators and students. “Google is an outsized player in both the search space and the education space.”

The company spearheaded a 1:1 device movement that put personal devices like Chromebooks into schools nationwide. Google Workspace for Education now powers administrative systems and classroom resources. Google later became a major evangelist for AI, introducing and then rescaling Gemini for Education. Today, millions of users from Kindergartners to post-grads rely on Google’s educational products.

By making generative AI fundamental to Search , to the point where many users don’t realize they’re using it , Google has eroded years of trust, sources told Mashable. Research shows that AI Overviews are prone to repeating misinformation, a problem that remains unresolved. Much like existing chatbot issues, the product’s design projects an air of authority that can confuse adults, let alone children.

Common Sense Media also found that Google’s AI tools readily completed homework assignments for minor users. Experts warn that the combination of misleading design, varying responses, and the agreeable nature of chatbots may lead students to unknowingly cheat. At minimum, searching on Google may be making learning harder.

Amanda Bickerstaff, deputy director of AI for Education, said, “They are infusing AI in developmentally inappropriate ways through Search.” Her organization works with schools to build AI literacy programs and reduce the negative impact of generative AI on children’s learning. “There is no reason why a 10-year-old, 13-year-old, or even 15-year-old should see an AI Overview or have AI Mode,” Bickerstaff added. “When teaching AI literacy, the first step is to be intentional. These modes are the opposite. You don’t go to a search engine to get a chatbot answer, you go to search for real things.”

According to the report, AI Overviews and AI Mode could also pose mental health and safety risks to children. Google does not offer the option to disable only AI features while keeping Search accessible. That means children can still be exposed to generative AI at school, even when guardrails are in place, simply by doing their schoolwork. Parents can turn off Search entirely or block Google.com for child accounts they manage. School administrators could theoretically do the same for Workspace for Education accounts, but Bickerstaff sees this as putting educators between a rock and a hard place. “What are you going to say? That kids can’t search the web?”

In a statement following the report’s release, Google said Common Sense Media’s stress testing was narrow and ambiguous, and not an “effective way to measure product safety and helpfulness.” The company noted that Common Sense Media tested only consumer accounts, not Workspace for Education accounts. However, Google confirmed that the AI-powered Search experience is the same for all users.

Because of this, some sources say it may be time to cut ties with Google. David Monahan, campaign director for child safety nonprofit Fairplay for Kids, stated, “Google’s AI Overview and AI Mode are undermining kids’ education, supplying them with dangerous disinformation, and putting their very lives at risk. If Google is not even willing to disable AI features for students, schools should stop using Google products altogether.”

The report could be a startling wake-up call for school districts, said Anya Meksin, deputy director of Schools Beyond Screens, a national coalition pushing for policies that limit tech use in classrooms. Schools Beyond Screens recently helped implement new tech use policies at Los Angeles Unified School District (LAUSD) schools, including a screen time ban for kindergarten and first grade classes.

Both Fairplay and Schools Beyond Screens are leading calls for years-long moratoriums on generative AI use in schools. Meksin explained that advocates have already sounded the alarm about AI Search and other integrated Google products. “We are shocked by the way Big Tech has colonized public education with their products,” Meksin said. “We have to treat this as an emergency. Education is the foundation of our democracy, and they are knowingly disabling our children from being able to fight back against these companies in the future.”

School districts, including LAUSD, have struggled to address problems posed by reliance on classroom tech, often in violation of their own burgeoning AI policies. “School districts are instantly in violation of any policies put in place because of the speed of the technology and the way AI is embedded,” Meksin said. For example, LAUSD’s new policy bans students under 13 from using generative AI in school. But through products like Chromebooks and Google Search on school computers, students can still interact with AI Mode.

Over the last 20 years, K-12 schools have spent billions to outfit classrooms with educational technology. Meksin said several districts told her it would be impossible to fully disable such expansive services, including Google products like Chromebooks. “We find this position to be completely laughable,” Meksin said. “School districts are liable for exposing children to products known to cause severe and lasting harm.” She believes educational institutions could enter the same legal hot water as social media and AI companies.

Google and its parent company Alphabet recently faced its first wrongful death lawsuit after a user died by suicide following conversations with Gemini. YouTube, also owned by Alphabet, was found guilty of negligent platform design and subsequent mental health harm to minor users. Torney likened the rollout of Google’s AI features to the introduction of YouTube to classrooms, which has also faced criticism from child safety advocates.

Instead of opening themselves up to greater liability, Meksin believes school districts should push for change from Big Tech, starting with stronger controls for embedded generative AI features now available on school devices. “As the largest purchasers of these products, districts have power. If Google refuses to comply, districts must cease using these products and void these contracts.”