Google said on Tuesday that it will switch off artificial intelligence features in its educational apps for K-12 student accounts unless school administrators explicitly enable them, and that schools which block the tools will also lose access to Google Search.

The company framed the first part as a concession to schools. After surveying administrators and teachers, it said, schools should control whether students use AI in the classroom. "We believe schools should have control over AI in their classrooms and in providing administrators with controls to help them manage their approach," Google said in a blog post.

The second part is where the argument started. If a school has banned the use of AI by students and teachers, Chris Phillips, Google's vice president for education, told The New York Times, it "will not have access to search, and they'll see a message from their school that indicates it's not available."

Why Search is now an AI product

Google's explanation is technical: AI capabilities are now deeply integrated into its search engine, so switching off AI for a student account and leaving search on is not a meaningful option.

Critics see the same fact from the other direction. If search cannot be separated from AI, then a school that wants to restrict AI is being asked to give up the most widely used research tool in education. Some called the announcement a "hostile ultimatum", a phrase that captures the structure of the choice even if it overstates the tone.

The announcement also included new AI products for classrooms and more controls for teachers, which the company presented as evidence that it is trying to accommodate concerns rather than override them.

The policy problem

Schools have spent two years trying to write rules for AI in classrooms, and most of those rules have been about what students may do with chatbots: whether they can use them for homework, how to cite them, and when their use counts as cheating.

Blocking a vendor's AI features is a different kind of decision, because it is about the infrastructure a school runs on rather than the behaviour of individual students. Google's Classroom and its education accounts are used by a large share of schools in the United States, which gives the company leverage that a smaller vendor would not have.

That leverage is the reason the announcement landed as a debate about market power rather than pedagogy. A school that wants to keep students off generative AI now has to weigh that against losing access to the search tools its teachers use every day.

What schools can do

In practice, schools have three options: enable the AI features with restrictions, block them and lose search, or move to another ecosystem. The third is expensive and disruptive, which is why most will probably choose the first, even if they would rather not.

The company has said the controls will let administrators decide which AI capabilities students can use, which suggests a middle path exists for schools willing to engage with the settings. For schools that have already banned AI outright, the choice is starker.

What happens next

Education departments and districts are likely to ask Google for a way to keep search while restricting AI, and regulators may take an interest in whether tying two products together is a fair practice. Google has not said whether it will offer such an option.

For parents and teachers, the practical question is what students will see in class next term. For Google, the question is whether a policy designed to give schools control ends up being remembered as the moment it used search access to keep them using its AI.

The procurement dimension is what makes the decision consequential. Schools rarely choose software one product at a time; they buy ecosystems, and switching costs are measured in training, integration and contracts that run for years. That is why a rule that applies to search access is more powerful than a rule that applies to a chatbot: it reaches every classroom that uses the platform for research, whether or not the teacher has any interest in AI. Districts that want to resist will need technical workarounds, legal advice and, in some cases, a different vendor entirely. The timing is also awkward for the company: the announcement landed in the same week as a national election that has made tech regulation a live political issue, which gives lawmakers an easy example to cite when they argue that platform power needs limits. Whether any of them act on it is another question, but the case for separating search from AI now has a concrete example attached to it.

Schools have been dealing with a different kind of disruption this autumn, from the hoax threat emails sent to UK schools to the consumer devices students use, such as the iPhone 18 Pro's variable aperture. The same week brought a reported plot targeting Nvidia's Jensen Huang, another reminder of how AI policy and security now overlap.