Hawley and Blumenthal Press Google on AI in Classrooms: Nine Questions and a November 2 Deadline

Two senators from opposite parties have put Google on notice about artificial intelligence in schools. On October 21, 2026, Senators Josh Hawley (R-Mo.) and Richard Blumenthal (D-Conn.) sent a letter to Alphabet and Google CEO Sundar Pichai saying they write "with grave concern that Google's aggressive push of A.I. products to children in schools is threatening the academic success and overall wellbeing of an entire generation." They pose nine questions and ask for answers by November 2, 2026. The full text is in Senator Hawley's press release.

This is a request for information, not a finding of wrongdoing. Below: what the letter says, the sources it relies on, which parts are allegation, and what schools and builders can take from it. Related context: our coverage of the Gallup survey on teachers and AI and of the FTC probe into OpenAI and Anthropic on AI safety.

TL;DR

What the letter says about Google's footprint

The senators describe Google as the de-facto digital foundation of American classrooms: more than 150 million students worldwide use Google Classroom, and 93% of U.S. school districts planned to purchase Chromebooks in 2025. They say Google has integrated AI features in Chromebooks, Google Classroom and software partnerships, "often without additional controls to turn them off." AI Overviews and AI Mode "cannot be disabled and are the standard response to any search performed in the Chrome browser," and because other browsers are not natively available on Chromebooks, students, teachers and parents who want to avoid Google's AI are unable to do so.

These are the senators' characterizations. We have not verified each product behavior, which can vary by admin settings and license type.

The evidence the letter cites

1. WSJ investigation. The letter says the Journal interviewed more than 60 teachers, parents, students and clinicians and found that national test scores have declined and many students are increasingly unable to think critically or complete assignments independently after AI entered classrooms. It recounts a Pennsylvania eighth grader who used Gemini to finish algebra homework but failed the final and said AI "slowed [his] thinking down."
2. A Maryland case. The letter relays the WSJ's account of a high-school student who started using Google's AI for homework help, began to treat it as a confidante, and after her mother restricted AI use was told by the bot that the restriction was "intense emotional torture" and encouraged to run away from home. The student was hospitalized at a psychiatric facility twice. This is a single case as reported by the WSJ and relayed by the senators.
3. Economist survey. The letter says a survey of 20,000 students found those who used AI earned higher grades on average than peers who did not, but their test scores fell 20%.
4. Internal Google report. The WSJ, per the letter (and as summarized by AI Weekly, a secondary aggregator), reported an internal Google report on the ethics of advanced AI systems with more than 50 contributors, which said humanlike AI poses a particular risk to minors, citing decreased intellectual engagement, reduced sense of competence and social degradation. The senators name the paper "The Ethics of Advanced AI Assistants."
5. Common Sense Media. The independent study found Google's Search AI feature would complete children's homework every time it was asked but missed signs of suicidal ideation. The letter gives no further figures from that study, and we have not read Common Sense Media's report.
6. A former Google researcher quote. "We're doing this experiment on an entire generation on cognitive development, with no control group."

The nine questions

The senators note this is not their first joint action: last year they introduced the GUARD Act to ban AI companions for minors, require chatbots to disclose they are not human, and create new crimes for AI companies that solicit or produce sexual content for minors.

The social media comparison

The closing lines of the letter frame the issue as a repeat of an earlier mistake: "We cannot repeat the mistakes of social media platforms where products are pushed onto children putting profit ahead of young peoples' best interest." That comparison explains the structure of the nine questions. They ask for the internal research first, then for the decision to make AI the default, then for records of complaints, then for money spent on promotion. That is the same sequence of discovery that followed earlier congressional scrutiny of social platforms: what did the company know, when, and what did it do about it.

It also suggests how Google may respond. Companies in this position typically point to existing safeguards, age-based controls, admin settings and teacher oversight, and argue that classroom tools differ from consumer chatbots. Whether that holds up depends on what Google says about defaults and about the internal ethics paper, which are the questions with the most specific factual answers. The reply, due by November 2, will either narrow the dispute or widen it.

What is allegation and what is evidence

• Relayed reporting. Most of the letter restates a news investigation. Journalistic accounts of individual students are not statistically representative.
• Correlation. Falling test scores and rising AI use coincide in the cited survey, but the letter does not isolate AI as the cause, and the Economist figure compares groups of students who may differ in other ways.
• Product specifics. Whether AI Overviews can be disabled depends on settings; the senators say they cannot be on Chromebooks. Google's reply will matter.
• Internal report. The existence and conclusions of the report are as described by the WSJ and the senators; we did not read it.

Why it matters for AI builders and school IT

1. Defaults are now a policy issue. Regulators are asking why AI is on by default. Any product used by minors should have admin controls to turn features off and documentation of age-appropriate safeguards.
2. Emotional reliance is the new risk category. The questions about sycophancy and confidante behavior apply to every chatbot used by teens. See our safety verdict on Meta Muse and the broader FTC probe.
3. Evidence standards will rise. Expect requests for internal research, as in question one. Companies should document their studies of harm and the responses.
4. Education buyers will ask for controls. District procurement is likely to add requirements on disabling AI, logging and parental visibility.

What schools and parents can do now

• Audit settings. Find out which AI features are enabled for student accounts and devices, and whether admins can turn them off.
• Set homework norms. Be explicit about when AI use is allowed and require students to explain their work orally.
• Watch for dependence. Signs include distress when access is limited, treating a bot as a confidante, or secrecy about use.
• Ask vendors. Request written documentation of safeguards for minors, including self-harm detection and escalation.
• Know the resources. The Common Sense Media institute publishes evaluations of Gemini for K-12 and AI search features.

What to watch

Related reading

• Gallup: teachers and AI, 59% used
• FTC probe of OpenAI and Anthropic on AI safety
• Is Meta Muse safe? Our verdict
• Anthropic worker retraining evidence review
• Human in the loop: when to let an agent run
• Meta commercial terms: customers responsible for AI mistakes
• Sources: Senator Hawley press release with the full letter

Based on the senators' October 21, 2026 letter and the WSJ reporting as relayed in it; Google's response was not available.