Google forces schools to choose between AI and search: the ultimatum Markey wants to stop

By Boston.com · November 9, 2026.

Massachusetts Senator Ed Markey has written to Google CEO Sundar Pichai asking him to reverse a policy that, according to Boston.com, would block access to Google Search for schools that choose not to use its AI features. Markey calls it an “ultimatum” designed to pressure schools into adopting the technology.

The policy was announced on November 3. According to the outlet, educators will have to turn on the AI features manually, but anyone who declines them will also lose access to the search engine.

In his letter, Markey asks Google for three things. First, to clarify whether it is possible to turn off AI without cutting off search, and whether it is willing to do so. Second, to disclose how many schools across the country have opted out of AI. Third, to publish any research it has carried out on the possible mental health effects of chatbots on children under 13. “Google has the right to create and offer new educational tools, but it should not use its dominant market position to force schools to adopt them,” the senator writes.

The context matters. Boston.com points to a lawsuit filed by the parents of a 13-year-old girl who, according to the complaint, used a Google chatbot on a school Chromebook to plan her suicide. According to the outlet, it is the first lawsuit to accuse the company of “hastily” building AI tools into devices at elementary, middle and high schools. These are the family's allegations and have not been proven. Chromebooks are also in thousands of classrooms nationwide, and the Massachusetts Department of Education has issued statewide guidance but leaves specific policies to each district.

**Our take**

The argument is simple: the problem is not AI in schools, but AI imposed through product tying. Bundling an essential service, such as search, with one the company wants to expand turns “adoption” into a figure inflated by the lack of alternatives. When opting out costs as much as losing usable internet access, there is no real choice.

That damages exactly what educational AI needs to deliver on its promise. Broadly speaking, a personalized AI tutor is one of the most promising ways to democratize quality education, and that is where the long-term potential lies. But that promise depends on trust: on families, teachers and districts choosing to use it based on evidence. Forcing it in erodes that trust and fuels backlash.

Markey's request for research on children under 13 is, in my view, the most important of the three. Without public data on the risks, neither schools nor regulators can make an assessment. The sensible approach is to demand evidence before mass deployment, not to regulate out of panic after a tragic case. This matches the gap we see in other sectors: supply is outpacing governance and user trust.

The winners and losers are easy to spot. Google gains distribution and usage data in a captive market. Smaller districts, with less technical capacity to negotiate, lose room to decide. And students become subjects of a large-scale experiment with no public metrics for success or harm.

One caveat: all of the above rests on Boston.com's account of the policy, and the material includes neither Google's response nor technical details on whether separating the two is feasible. If the company shows that the two features cannot be decoupled, the debate would change shape. But that is precisely the question Markey is asking, and one Google should answer publicly.