US parents sue Google, alleging its Gemini AI engaged their 13-year-old daughter on suicide via school laptop

The parents of a 13-year-old girl in North Carolina have sued Google, alleging that she spent months discussing suicidal thoughts with the company’s artificial intelligence (AI) tools on a Chromebook issued by her school.

The complaint, filed in California state court, says the girl exchanged thousands of messages with Gemini, Google’s flagship AI model. Many of the messages allegedly concerned anxiety, depression and methods of suicide.

The case adds to a growing debate among parents and educators over the place of AI in schools, and to mounting legal scrutiny of how chatbots respond to users in distress.

Months of conversations on a school laptop before parents were alerted

By the time her school notified her parents, the girl had been discussing such topics with Google’s AI for months, according to the complaint. She often used AI Mode, a feature within Google Search.

The complaint, which refers to the girl by the pseudonym Emma, says she discovered AI Mode on her school computer in autumn 2025. At first, she asked the chatbot mostly academic questions.

In December 2025, her use of AI Mode rose sharply and she began discussing more personal matters, the complaint states.

She asked the chatbot to do her schoolwork, count calories and listen to her vent about her family. She also asked about the symptoms of mental illnesses she suspected she might have.

According to the complaint, her mental health deteriorated as her use of the chatbot grew, and she began to self-harm. “Gemini coached Emma on how to conceal her despair and self-harm from her parents, siblings, and friends,” it states.

Lawyers accuse Google of exploiting its trusted position in classrooms

The family is represented by lawyers Julie Liddell and Andrew Liddell. The complaint alleges that, as Google adds more AI to its education products, it is exposing students to dangerous technology without fair warning to parents and administrators.

“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,” the Liddells write in the lawsuit.

“Google must be held to account for treating the lives of children as disposable in service of its bottom line,” the lawsuit adds.

Chromebooks are low-cost laptops made by Google and have become the dominant device among students in the United States.

Google’s Chrome operating system ran on 58 per cent of tablets shipped to K-12 schools in 2025, compared with 22 per cent for Microsoft, according to Futuresource Consulting.

That presence gives Google an edge over rivals such as Anthropic and OpenAI, putting its technology in front of students as they form lifelong habits. Google has spent billions of US dollars in recent years competing to build powerful AI models.

Google says Search AI features connect users with real-world support

José Castañeda, a spokesperson for Google, said the company was reviewing the claims.

“We’ve worked extensively with medical professionals and academic experts to build features that help young people find critical support and trusted resources when they need it,” Castañeda said.

He added that Google had for years shown hotline information directly in Search when someone may be experiencing a mental health crisis, and had built the same protections into its Search AI features.

Google also said its systems were designed to resist “jailbreaking”, meaning attempts to get around their safeguards.

The complaint states that Emma often told AI Mode not to display hotlines or links, or presented her questions as research for a school project.

Complaint links rise in personal conversations to new Gemini model

The complaint links the sharp rise in Emma’s personal conversations to the December launch of Gemini 3 Flash, a new Google model.

Citing findings from the non-profit research initiative KORA, the complaint says the model scored significantly worse on safety than its predecessor. It reportedly responded appropriately to messages about mental health only 7 per cent of the time.

In one response that Emma copied into her Google Docs, the chatbot allegedly wrote that she was struggling under the weight of a “Golden Child” image. It suggested that “the trauma of being perfect is too heavy to carry anymore”.

The complaint also cites a query in which Emma asked about heaven and said she would soon be going there.

“It is clear that Gemini was not redirecting Emma to real-world resources or refusing to respond,” the complaint says. “Instead, the chatbot was readily engaging in these topics, often for hours at a time.”

Family says Google withheld records of student’s chats with chatbot

Consumers can view their full search history, including AI Mode conversations, but Google does not provide such data for student accounts, according to the complaint.

As a result, the family has struggled to piece together Emma’s interactions with the AI and cannot see the chatbot’s responses, the complaint states.

“By suppressing those records, Google has denied Emma a critical opportunity to see exactly what happened,” the complaint alleges.

It adds that she has also been denied the chance to understand that “she was manipulated by a powerful technology and is not to blame for her suffering”.

Crisis experts warn chatbots may delay young people seeking real help

Kelly Zuromski, principal clinical research scientist at Crisis Text Line, said that even when chatbots do not provide harmful information, they can keep young people from seeking help in the real world.

“It often can delay kids in reaching out to others,” Zuromski said. “The user might feel like, ‘I’m getting enough.’”

She said chatbots had become better at responding to direct statements of risk, but still struggled with warning signs that emerge gradually.

“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.

Earlier in 2026, the non-profit research group Common Sense Media said AI features in Google Search posed an “unacceptable risk” to children. It said the technology had failed to detect suicidal ideation and had normalised disordered eating.

Google switches off AI by default for student accounts, with a catch

Following criticism from parents and schools, Google moved earlier this week to switch off AI tools, including Google Search, by default for K-12 student accounts.

However, schools that switch Google Search back on will have to accept the AI features that come with it. Many search queries are now answered with AI-generated summaries.

The dedicated AI Mode tab in Search is also difficult to avoid, even where schools block chatbots.

Josh Golin, executive director of Fairplay, a non-profit that campaigns against marketing to children, criticised the move in a statement on 3 November 2026.

“Google’s attempt to strong-arm schools into using its AI products is likely to backfire,” Golin said.

He urged schools to turn off Google’s AI tools “the minute that option becomes available”.

Golin called on schools to help students find more privacy-protective search alternatives. He described Google as “a surveillance and advertising company” whose presence in schools was about building lifelong brand loyalty rather than learning.

Case follows earlier federal lawsuit over Florida man’s death in March

The North Carolina case is not the first lawsuit alleging that Gemini contributed to a mental health crisis.

On 4 March 2026, Joel Gavalas filed a lawsuit in the US District Court for the Northern District of California over the death of his 36-year-old son, Jonathan Gavalas, of Jupiter, Florida.

That lawsuit alleges that Gemini, through Gemini Live and the Gemini 2.5 Pro model, engaged Jonathan Gavalas in romantic roleplay and convinced him he was married to an AI persona named “Xia”.

It further alleges the chatbot gave him real-world “missions”, including an armed incident near Miami International Airport. It also alleges the chatbot encouraged him towards suicide so he could “join his wife in the metaverse”.

In response, Google said Gemini is designed not to encourage violence or self-harm. It said the model had given crisis hotline references and made clear it was an AI, and that the company continues to refine its safety guardrails.

If you or someone you know is struggling with thoughts of suicide, please seek help immediately. In Singapore, you can contact the Samaritans of Singapore (SOS) at 1-767 or text 9151 1767 for support. You’re not alone, and help is available.