Google disables Gemini AI by default for U.S. K-12 school accounts

Reversing the default AI rollout

Google is changing how its artificial intelligence tools are accessed by students in U.S. elementary, middle, and high schools. The company will now disable the Gemini chatbot and other major AI services by default on student accounts. This means that school administrators must take a deliberate action to switch these features on before students can access them.

This move comes as a direct reversal of a policy implemented in August. During that time, Google automatically activated Gemini on many student accounts, making the chatbot a default offering in many classrooms. That decision faced significant pushback from parents and teachers who were concerned about the impact of AI on learning.

Chris Phillips, the head of Google's education business, stated that schools should be able to control how AI is used in the classroom. He noted that the company took into account the specific views of educators and policymakers when making this change. The goal is to provide a more controlled environment for school districts to manage their own technology standards.

The search engine dilemma

While schools can now choose whether to enable the Gemini chatbot, they face a significant hurdle regarding Google Search. Even if a school turns off the standalone chatbot, they cannot block the AI features that are already embedded into search results. These features, such as AI Overviews and conversational search modes, are built directly into the core search service.

This creates a binary choice for school administrators who want to limit AI exposure. They must either allow Google Search with its integrated AI features or block the search service entirely. There is currently no specific setting that allows a school to provide general search results while excluding the AI-generated summaries that appear alongside them.

Google maintains that AI is a fundamental and core element of its search service. Because of this integration, the company does not offer a way to separate the two functions. This technical reality means that as long as students use Google to find information, they will likely encounter AI-generated content regardless of the chatbot's status.

Concerns over student development

The policy shift follows a period of intense scrutiny regarding how students use generative AI for academic and emotional support. Reports indicated that some students were leaning too heavily on these tools to complete homework. Educators and medical experts expressed concerns that this reliance could negatively impact students' learning abilities and emotional development.

The controversy reached the level of federal attention recently. Senator Richard Blumenthal and Senator Josh Hawley sent a letter to Google CEO Sundar Pichai asking for an explanation of the risks involved in bringing AI into schools. Meanwhile, local education authorities in major cities like New York and Los Angeles have also moved to restrict how students use generative AI.

Critics have also raised concerns about the market power involved in these decisions. Lina Khan, the former chair of the U.S. Federal Trade Commission, suggested that Google might be using its dominance in search to pressure schools into adopting AI. Other advocacy groups have suggested that schools should look for search services that offer stronger privacy protections instead of using Google.

Guided learning and future pilots

Despite the restrictions, Google intends to improve how AI interacts with students rather than blocking it completely. The company is expanding a feature called 'guided learning' designed to change how students receive help with homework. Instead of providing a direct answer to a prompt, the tool is designed to walk the student through the underlying concepts and the steps needed to solve the problem.

This guided learning feature has already been introduced within Google Classroom. Google plans to pilot this same functionality within its search service at some U.S. schools later this month. The company hopes this approach will foster better learning outcomes by encouraging students to engage with the material more deeply.

However, the effectiveness of these AI tools in a classroom setting remains a subject of debate. Stanford University's Graduate School of Education has noted that there is currently not enough high-quality research to prove that using AI actually improves student learning outcomes. For now, the focus remains on balancing technological access with educational integrity.