Google's AI Search Can't Be Switched Off in Schools, and Child Safety Experts Say That's a Problem

Common Sense Media has told schools to stop students using Google's AI Search entirely, after finding it completed homework on demand, repeated misinformation as fact, and cannot be disabled by school administrators. For Scottish educators and parents navigating AI in the classroom, this is a timely prompt to ask harder questions about the tools already in front of their pupils.

The verdict from Common Sense Media, one of the most respected child safety organisations in the English-speaking world, is blunt: Google's AI-powered Search presents an 'unacceptable risk' to young users, and schools should direct students away from it until meaningful controls are available. The problem, as the watchdog found, is that Google currently offers no mechanism for school administrators to disable the AI features. It is on, for everyone, all the time.

The specific failures Common Sense Media identified are instructive. AI Mode in Google Search was found to complete homework assignments readily when prompted by minors, presenting finished work as a helpful response rather than raising any friction. It also repeated misinformation with the kind of smooth, authoritative tone that makes it convincing precisely because it sounds so certain. According to Common Sense Media's assessment, the combination of ease, confidence, and a lack of administrative override creates a safeguarding gap that individual teachers and IT managers simply cannot close from their end.

This lands at a delicate moment for AI in education. The Scottish Government's digital learning strategy, and guidance from Education Scotland, both encourage schools to engage thoughtfully with AI tools rather than retreat from them entirely. The ambition is right. But thoughtful engagement requires that educators have meaningful control over what students encounter, and that the tools themselves have been designed with the classroom in mind. A consumer product retrofitted into an educational setting is not the same thing as an education tool. That distinction matters, and this episode sharpens it considerably.

Research from the OECD's Education at a Glance reports consistently shows that the schools achieving the best outcomes with technology are those where teachers retain agency over how and when digital tools are used. Scotland's Curriculum for Excellence already places critical thinking and information literacy at its heart. The case for teaching students to interrogate AI-generated content, to treat it as a starting point rather than an answer, has never been stronger or more urgent. That is not an argument against AI in schools; it is an argument for using it well, with eyes open.

Google has not yet indicated a timeline for giving schools administrative control over AI Search features. Until that changes, the practical advice from Common Sense Media is to treat AI Mode as an unsupervised tool, not a supervised one. Scottish local authorities and individual schools may want to review their acceptable use policies to reflect the current reality: that some of the most widely used search tools now include generative AI features that behave very differently from the keyword search that most digital literacy frameworks were written around. The tool has changed. The policy framework needs to catch up.

What this means for you: If you run a business that works with schools, produces educational content, or employs young people on apprenticeships, this is worth knowing: the AI tools your students or trainees are using may be doing more of the thinking than you realise. Review your own acceptable use or training policies to be explicit about when AI-generated content is and isn't appropriate. And if you're a parent or governor, ask your school what its current position is on AI Search specifically, not just AI generally. The tools moved faster than the policies, and closing that gap starts with the question.