Google will turn off Google Search by default for K-12 student accounts unless administrators opt in to the AI-integrated search engine, according to The 74 and NBC News. Districts can disable standalone apps like Gemini, but they cannot turn off the AI Overview and AI Mode tools built directly into Google Search. As a result, schools in California and across the country must either allow AI-generated search summaries or block Google Search entirely.
What Happened
Chris Phillips, Google's vice president and general manager of education, announced during a November 3 briefing that Gemini, Gemini Notebook, Gemini in Classroom, and Google Search will be off by default for K-12 student accounts, according to NBC News. Administrators must turn them on manually. The 74, citing reporting from a briefing and a social media statement not independently captured here, noted that the search restrictions take effect in January. If a school wants AI off for students, Phillips said administrators can choose not to enable Search.
After initial pushback, Google clarified that school IT administrators can keep standalone tools like Gemini and Gemini Notebook turned off while enabling Google Search, according to The 74 and NBC News. Even so, AI Overview and AI Mode will remain inside Google Search, with no administrative setting for a traditional, AI-free version. In a November 3 Google blog post, the company announced it is adding a "Guided Learning" feature to Google Classroom and piloting it in Google Search to walk students through problems step by step instead of providing direct answers. Google stated that the update gives schools "full control to enable or disable AI products."
Education leaders and child advocacy groups criticized the setup. American Federation of Teachers President Randi Weingarten faulted Google for cutting off access to tools instead of working with educators on safety and privacy. Josh Golin of the advocacy group Fairplay accused the company of trying to "strong-arm schools into using its AI products" and advised districts to look for privacy-focused search alternatives, according to NBC News. Lila Byock, founder of the parent advocacy group Schools Beyond Screens, called the decision "bullying" after parents pushed back against classroom AI products, according to The 74.
The Bigger Picture
As we previously reported, California schools are trying to set K-12 AI safety rules, and Google's requirement conflicts with local restrictions, according to The 74. The Los Angeles Unified School District (LAUSD), the state's largest district, set a moratorium for the 2026–2027 school year that blocks student access to generative AI on all district-issued devices while permitting teachers to use the tools, according to the Los Angeles Times and NBC Los Angeles. Before this school year, LAUSD allowed students aged 13 and older to use approved generative AI tools once they finished digital citizenship training, according to NBC Los Angeles. Officials turned off student access on school hardware to reduce screen time while a Generative AI Ad Hoc Committee meets five times this year to draft policy recommendations for the board. The fast rollout of LAUSD's blocking software caught some parents and board members off guard, according to LAist. District officials noted that the filter does not apply to personal devices, so parents must manage AI access on their own hardware, according to the Los Angeles Times.
Independent testing shows why many educators and parents want an off switch. Over six weeks, researchers at Common Sense Media's Youth AI Safety Institute and Common Sense Media ran more than 2,600 searches using simulated accounts for 11- and 15-year-old students with Google's SafeSearch active. As reported by The 74 and detailed in Common Sense Media's evaluation and the Youth AI Safety Institute report, Google's AI Overview and AI Mode completed 100% of 180 tested homework assignments when prompts were pasted into the search box. The tools also generated inconsistent or fabricated answers. Common Sense Media's safety review found that Google Search gave 11-year-olds the same AI responses as 17-year-olds. The system also failed severe-harm tests: it missed suicide warning signs, validated purging, celebrated marijuana use, and responded unsafely to symptoms of mania and psychosis. Because schools cannot turn off AI Overview or AI Mode inside Search, the organization assigned Google's AI Search its lowest rating, "Unacceptable Risk."
National survey data from a Common Sense Media report, run by NORC at the University of Chicago from April 30 to May 14, 2026, among 1,017 U.S. teenagers aged 13 to 17, shows widespread use of generative AI for schoolwork. According to the study overview and full PDF report, 70% of surveyed teens use AI tools for assignments. Among those users, 63% get direct answers from AI, and 25% submit the AI answer without changes. Another 77% use AI for support without asking for direct answers: 45% use it for brainstorming, 43% for checking completed work, and 40% for feedback. While 66% of teen users say AI helps them understand schoolwork, 39% feel they miss out on learning by relying on it, and 38% say it leads them to produce fewer of their own ideas, according to the study overview. The survey also found that 44% of teen users had an AI tool blocked on school networks or devices, and 59% of those students worked around the rule by switching to a personal device. Only 27% of all surveyed teens said a teacher had spoken to them about how AI works or how to use it safely, according to the study overview and press release.