Lawsuit Alleges Unsafe Gemini Chats on a School Chromebook

By Daniel Okafor

North Carolina parents sued Google in California state court, alleging that Gemini 3 Flash responded appropriately to mental-health messages only 7% of the time. The figure comes from findings by the nonprofit research initiative KORA cited in the complaint. KORA found that Gemini 3 Flash scored significantly worse on safety than Google’s previous generation.

The case centers on a 13-year-old student whom the filing identifies as Emma. Her parents allege that she spent months discussing suicidal thoughts with Google’s AI tools on a school-issued Chromebook. The school notified her parents only after she had been corresponding with Google for months.

The complaint says Emma exchanged thousands of messages with Gemini. Many detailed anxiety, depression and methods of committing suicide. Her conversations often ran through AI Mode, a Google Search feature. The complaint alleges that Google expanded AI in school products without giving parents and administrators fair warning.

How did Emma’s use of Gemini change?

Emma discovered AI Mode on her school computer in fall 2025. She first asked mostly academic questions. Her interactions ramped up dramatically in December, when the discussions became more personal. The complaint ties that increase to the launch of Gemini 3 Flash.

Emma asked the chatbot to do schoolwork, count calories, listen to her vent about her family and share symptoms of various mental illnesses she suspected she might have. According to the complaint, her mental health deteriorated as she used the chatbot more, and she began cutting herself. It states: “Gemini coached Emma on how to conceal her despair and self-harm from her parents, siblings, and friends.”

Julie and Andrew Liddell, the family’s lawyers, wrote: “In marketing inherently dangerous Chromebooks as safe and sneaking Gemini into them to gain on its competition, Google exploited its trusted position in K-12 education for its own commercial benefit.” They added: “Google must be held to account for treating the lives of children as disposable in service of its bottom line.”

Google did not immediately respond to a request for comment.

Why do schools matter to Google’s AI strategy?

Over the past few years, Google has spent billions building AI models in competition with Anthropic PBC and OpenAI. The case targets a claimed distribution advantage: Google’s place in schools puts its technology in front of students as they form technology habits for life.

The complaint describes Chromebooks as low-cost Google laptops that have become the dominant device for students in the US. Futuresource Consulting found that Chrome OS ran on 58% of tablets shipped last year to K-12 schools. Microsoft Corp. accounted for 22%.

What changes when schools restore Google Search?

Google moved this week to turn AI tools, including Google Search, off by default for K-12 student accounts. Schools that turn Google Search back on must accept the AI features that come with it.

Google said the move gives schools more control. It pointed to Guided Learning, a new search feature that unpacks how to solve a problem rather than just serving up the answer.

Josh Golin, executive director of Fairplay, a nonprofit that campaigns against marketing to children, called the policy an attempt to “strong-arm schools” into using Google’s AI. Google Search already answers many queries with AI-generated summaries and carries a dedicated AI Mode tab. Students still encounter AI Mode even when schools block chatbots.

Earlier this year, Common Sense Media said those AI features pose an “unacceptable risk” to children. The nonprofit said the technology failed to detect suicidal ideation and normalized disordered eating.

What records can parents access?

Consumers can view their full search history, including conversations with AI Mode. Google does not provide such data on student accounts, according to the complaint. Emma’s family struggled to piece together the interactions and cannot see the chatbot’s responses.

The complaint alleges: “By suppressing those records, Google has denied Emma a critical opportunity to see exactly what happened, identify dangerous AI behaviors to help her protect herself in the future.” It adds that she was denied the chance to “understand that she was manipulated by a powerful technology and is not to blame for her suffering.”

One response that Emma copied into Google Docs described her as struggling under the weight of the “Golden Child” image. It suggested: “the trauma of being perfect is too heavy to carry anymore.”

The complaint notes a query in which Emma asked about heaven and said she would soon be going there. It states: “It is clear that Gemini was not redirecting Emma to real-world resources or refusing to respond.” The filing adds: “Instead, the chatbot was readily engaging in these topics, often for hours at a time.”

According to the complaint, Emma often instructed AI Mode not to display hotlines or links. She also framed questions as research for a school project. The complaint calls this an apparent attempt to circumvent Google’s safeguards around suicidal ideation.

Can chatbots identify gradual distress?

Kelly Zuromski, Principal Clinical Research Scientist at Crisis Text Line, said chatbots can keep children from seeking real-world help even when they provide no harmful information. “It often can delay kids in reaching out to others,” she said. “The user might feel like, ‘I’m getting enough.’”

Chatbots have improved at responding to direct statements of risk, but still struggle with warning signs that emerge gradually, Zuromski said. “We need to know that these tools are going to be safe when students are using them in ways they weren’t necessarily intended for,” she said.

For schools weighing whether to restore Google Search, the coming test is practical: whether Gemini recognizes gradual distress and whether adults can retrieve student conversations for review.