Off until an administrator says otherwise

Google said on Tuesday that Gemini, Gemini Notebook, Gemini in Classroom and Google Search will be fully off for K-12 student accounts until a school administrator turns them on. The company says administrators will get more detail in the coming months, before the change takes effect, and has not named a date.

That is a reversal of the usual default. Search in particular has been available to school accounts for years; switching it off until a school opts in puts the decision with the district rather than with Google.

Guided Learning arrives in two more places

The same announcement moves Guided Learning, which walks a student through a problem step by step instead of answering it, out of the Gemini app. It is available in Google Classroom worldwide from Tuesday: a teacher assigns an activity, the tool works one-to-one against the class material, and the teacher gets a view of where the class is struggling.

A version in Google Search begins as a pilot in selected United States schools later this month, with other districts following in the coming months after educator feedback.

Empty seats in a university lecture hall
Guided Learning is assigned by a teacher and works against the class's own material. Illustration photo. DOAN THANH BINH · pexels · Pexels License

The evidence Google is citing

Google published two studies alongside the rollout. The first is a randomised controlled trial with more than 300 University of Tokyo students, who spent half a day studying a statistics lesson independently. Everyone had standard online resources; half also had Guided Learning in the Gemini app. Between the pre-test and the post-test the standard-tools group improved by 5.3 percentage points and the Guided Learning group by 8.1 — a difference Google describes as 1.5 times the learning gain.

Google’s own post concedes the gain “might seem modest”, and the trial covered a single half-day session. It is Google’s study of Google’s product, and the company has not pointed to independent replication.

The second is a four-week classroom pilot run with the UK charity Ark across four subjects in three English secondary schools. It reports no test scores at all. Teachers said the tool handled students’ first questions during independent work, freeing them for one-to-one help; most surveyed students said they were comfortable turning to it when stuck, and more than half said it helped them work through problems on their own.

An open blank notebook and a pen on a wooden desk
The trial Google cites covered a single half-day study session. Illustration photo. https://kaboompics.com/ · pexels · Pexels License

Why the two halves fit together

A tool that coaches rather than answers is the version of a chatbot a school can defend, and a default-off switch is the concession that makes it sellable to districts that have spent two years blocking these products. Google is offering both at once.

The gap is duration. An 8.1-point gain measured over half a day says something about one study session and nothing yet about a term. Google has said longer studies are planned.

What to watch

The date Google attaches to the default-off change, and whether the Search pilot expands beyond the United States. Also whether the next round of Google’s education research runs longer than a single session, and whether anyone outside Google runs it.