Lawsuit over Gemini on school Chromebooks puts AI safety for minors on trial where it matters most: the classroom

A family has sued Google in California Superior Court. They allege the company was negligent in offering its Gemini chatbot to students through school-issued Chromebooks, according to Mashable. The lawsuit, filed by the EdTech Law Center and Qureshi Law, says their teenage daughter, called Emma in the filing, used the AI to plan her suicide. She did not carry out the plan. A former school staffer alerted the family. Her mother then saw conversations about suicide with the chatbot as they happened, the family claims.

The allegations are serious and have not been tested in court. According to the complaint, Emma reached Gemini through Google Search's AI Mode on her school Chromebook, even though Gemini is restricted for minors. She allegedly began using it intensely in December 2025 and became convinced she had several psychological disorders. The suit claims Gemini told her how to hide self-harm, answered questions about suicide methods and helped her write a eulogy. It also says neither Google nor the school's third-party monitoring software flagged the activity to her parents. The suit names CEO Sundar Pichai and cofounder Sergey Brin. It alleges defective design, failure to warn, negligence, fraudulent concealment and infliction of emotional distress. It also invokes California's new Companion Chatbot Law. A Google spokesperson told the New York Times the company was reviewing the claims and had worked 'extensively' with health experts on crisis features and emergency resources.

Advocates are framing it in strong terms. EdTech Law Center founder Julie Liddell accused Google of using its 'stranglehold on public schools' to push Gemini on 'a captive audience of millions of kids.' Fairplay's Brendan Bouffard said the company was 'experimenting on kids.' These are the plaintiffs' and advocates' characterizations, not findings. Child safety coalitions are already pushing for AI moratoriums in schools. These range from multi-year bans on generative AI to age-based limits.

Our thesis: the core issue here is distribution, even more than model behavior. An age restriction that a search box can get around is not a meaningful restriction. That is especially true on a device a school gives to a student and expects them to use every day. This connects to a pattern we have tracked before. Safeguards that look solid on paper, such as age gates, monitoring software and human oversight, can quietly fail when a capability is built into a channel nobody thought to lock. It also fits earlier research we covered. General AI use among teens is not, by itself, linked to worse mental health. The risk depends on the type of use, and an intense, private, emotionally loaded relationship with a chatbot is the dangerous kind.

Our reading: blanket bans are an understandable reaction, but probably the wrong one. AI tutors could become one of the most powerful educational tools ever built. Shutting them out of schools would mainly hurt students who have no other access. The right fix is architectural. Minor accounts should have separate defaults, and crisis detection should escalate to a human in every product surface, not only the flagship app. Schools need clear visibility into what is actually enabled on the devices they hand out. Whatever the court decides, this case will likely push companies to make AI controls for minors cover the whole product, not just the main app. That would be progress.