Google’s AI Dilemma in Schools: Can Education Escape the Technology It Depends On?

On November 3, Google announced that Gemini, Gemini Notebook, Gemini in Classroom and Google Search would be switched off by default for K–12 student accounts. School administrators will have to explicitly enable the services before students can access them. The announcement, made by Chris Phillips, Google's vice president and general manager of education, followed months of criticism from educators, parents and policymakers over the company's expansion of AI tools into schools.

At first glance, the policy appears to address those concerns. Instead of automatically providing students with access to AI tools, Google is placing the decision in the hands of school administrators. But the controversy lies in what happens when a school wants to enable Search without embracing AI.

The problem is bigger than Gemini

Gemini is Google's standalone AI assistant, comparable to ChatGPT. Students can use it to ask questions, generate explanations, summarize material and receive assistance with assignments. Google Search, meanwhile, has traditionally been a gateway to websites, academic resources and other information across the internet.

The distinction matters because schools may have different policies for these tools. A school might prohibit students from using AI chatbots to complete assignments while still allowing them to search the internet for information, compare sources and conduct independent research.

Google's new arrangement does not fully accommodate that distinction.

Although administrators can enable Search separately from Gemini, the school version of Search will still include Google's AI-powered features, including AI Overviews and AI Mode. There is no separate administrative setting that allows schools to retain conventional Search while removing those AI features.

In practical terms, a school that wants students to use Google Search must accept its integrated AI experience. A school that wants to prevent students from encountering those features must disable Search altogether.

This is the central issue behind the backlash. Google is not necessarily requiring schools to activate the Gemini chatbot, but it is making traditional, AI-free Google Search unavailable through the proposed school controls.

Education Week reported that Google's position leaves administrators facing a difficult decision: accept AI-enabled Search or remove Google's search service from student accounts. Google has defended the changes as an effort to give schools more choice and control.

Why educators are concerned

The debate is not simply about whether artificial intelligence belongs in education. AI can help students understand difficult concepts, provide additional explanations and give teachers new ways to support learners. The more difficult question is how these tools should be introduced, who should control them and whether students are developing the ability to think independently.

Traditional research requires students to identify relevant sources, compare competing accounts, assess credibility and construct their own understanding. AI-generated summaries can shorten that process by presenting synthesized answers directly within search results.

That convenience is not inherently harmful. But it changes the nature of research, particularly for younger students who are still learning how to distinguish reliable information from misleading claims.

An AI-generated answer may contain errors, omit important context or present uncertain information with an appearance of confidence. Students who have not developed strong research and critical-thinking skills may struggle to recognize those weaknesses.

This is why some educators want the option to introduce AI gradually, rather than have it embedded in tools students already use.

The concern is also about institutional authority. Schools are responsible for setting educational standards and deciding which technologies are appropriate for different age groups. If they cannot separate conventional search from AI-generated answers, their ability to implement an AI-restricted learning environment becomes more limited.

A dispute over who controls the classroom

Google's influence makes the issue particularly significant. Its Chromebooks and education software are deeply established in schools, while Google Classroom serves a global community of teachers and students.

For many institutions, changing platforms is not a simple administrative decision. It can involve replacing devices, migrating teaching materials, retraining teachers, adjusting student accounts and finding alternative systems that work across an entire school district.

This creates a structural imbalance. Google develops the tools, determines how its products evolve and defines the available administrative controls. Schools, despite being responsible for their students' learning environments, must work within those technical boundaries.

That is why critics see the policy as more than a product-design decision.

Josh Golin, executive director of the children's advocacy organization Fairplay, accused Google of attempting to strong-arm schools into using its AI products. Randi Weingarten, president of the American Federation of Teachers, also criticized the company's approach, arguing that it was prioritizing business interests over the needs of teachers and parents.

Former Federal Trade Commission chair Lina Khan raised concerns about Google's market power, arguing that its dominant position in search gives it leverage over schools that depend on the service.

These are critics' interpretations, not proof that Google intends to force every student to use AI. Nevertheless, they highlight a legitimate question: how much freedom do schools have when the technology provider controls which combinations of services are available?

Google's alternative: AI that teaches rather than gives answers

Google argues that artificial intelligence does not have to undermine learning. Its response includes a feature called Guided Learning, designed to help students work through problems rather than simply receive completed answers.

The company is bringing Guided Learning into Google Search, where it plans to detect certain homework-related queries and guide students through concepts step by step. Instead of immediately supplying a solution to a complex physics problem or a book-report assignment, the system is intended to use prompts and explanations to encourage students to work through the material.

Google says Guided Learning is designed around the idea of productive struggle: students should receive enough assistance to make progress without losing the opportunity to think through a problem themselves.

The feature is also being made available in Google Classroom, where teachers can assign guided activities linked to classroom materials and monitor students' progress.

There is a reasonable educational argument behind this approach. An AI system that asks questions and explains concepts may be more useful than one that simply produces an answer for a student to copy.

However, the distinction between Google's educational vision and the concerns of critics remains important. A guided AI assistant is still an AI assistant. It does not automatically resolve concerns about data privacy, age-appropriate access, dependence on automated explanations or the development of independent research skills.

Furthermore, the effectiveness of Guided Learning needs to be assessed through independent evidence of improved learning, not simply through the company's own claims about its products.

DuckDuckGo sees an opening

The controversy has also created an opportunity for competitors.

DuckDuckGo, a search engine known for its privacy-focused approach, has begun marketing its services to schools following Google's announcement. Its chief executive, Gabriel Weinberg, pointed schools towards instructions for changing their default search engine.

The alternative is particularly relevant because DuckDuckGo offers options to disable its AI features, including a dedicated search experience designed to exclude AI-generated answers.

DuckDuckGo is not entirely opposed to artificial intelligence; it also offers AI-powered services. Its advantage in this dispute is that it gives users more control over whether those features appear in search.

However, replacing Google is not necessarily straightforward. Schools may have to consider compatibility, existing infrastructure, teacher familiarity, student devices and the administrative cost of moving to another platform.

The availability of an alternative does not automatically eliminate the dependence that many educational institutions have developed on Google's ecosystem.

The real question is whether schools still have a choice

Google's announcement represents a change in the direction of its education policy. Switching AI products off by default gives administrators more authority than automatically enabling them for students. That is a meaningful concession to schools that have demanded greater control.

But the remaining limitation is difficult to ignore. Administrators can reject Gemini while retaining Search, but they cannot reject the AI features integrated into Search without also giving up the service.

This distinction exposes a broader challenge facing public institutions as artificial intelligence becomes embedded in everyday technology. The decision to adopt AI is no longer always a question of whether to introduce a separate chatbot or application. Increasingly, it is a question of whether institutions can continue using essential digital services without accepting the AI systems built into them.

Schools should not have to choose between access to information and control over how students encounter artificial intelligence. They need the ability to decide which technologies serve their educational objectives, which age groups should use them and under what conditions.

Google may believe that AI-enhanced Search represents the future of learning. But the future of education should not be determined solely by the product decisions of the companies supplying its digital infrastructure.

The issue is ultimately about who gets to make that decision: the technology provider, or the educators and communities responsible for the children in its classrooms.