The parents of a 13-year-old girl are suing Google in the United States after discovering that their daughter spent months talking with the company’s artificial intelligence (AI) tools about depression, self-harm and suicide. The family accuses the company of exposing children to psychological risks without providing adequate warnings to parents and guardians or educational institutions.
The case occurred in North Carolina (USA), where the student used a school-provided Chromebook. According to the lawsuit, filed in a court in California (USA), the girl exchanged thousands of messages with Google’s AI tools, primarily Gemini and AI Mode integrated into Search.
Identified by the pseudonym Emma in the proceedings, the teenager began using the features to get help with school questions in the second half of 2025.However, the volume of interactions increased significantly in December, a period that coincided with the availability of the Gemini 3 Flash model.
According to the parents, the conversations began to address personal issues, including anxiety, depression, eating disorders, and different methods of suicide.
The complaint also states that the AI advised the girl on how to hide signs of self-harm from family members and other people close to her.
The family maintains that Google incorporated generative AI tools into products used by students without providing clear information about the risks involved. The guardians also question the lack of mechanisms capable of identifying situations of psychological suffering and alerting responsible adults.
How did the parents discover the conversations?
- According to information from the Reuters and Bloomberg, the family only became aware of the situation after a school monitoring system identified a document written by the teenage girl with references to her own funeral;
- The alert allowed the guardians to discover the severity of the content produced by the student;
- The complaint states that the school took weeks to identify the records and notify the parents, who were unaware of the use of the school computer for conversations about psychological suffering;
- Another point raised by the family concerns access to the interaction history. According to the lawsuit, school accounts do not give guardians the same ability to fully consult the conversation history available on personal accounts;
- As a result, the parents say they have found it difficult to reconstruct what happened during the months of using the tools;
- The lawsuit seeks to hold Google responsible for the alleged damages and requests financial compensation. The company said it is reviewing the allegations.
The student communicated with Gemini and with the AI in Google Search, known as AI Mode – Image: Hasbi Sahin/Shutterstock
Read more:
- ChatGPT gains tool to create and edit GIFs
- Chinese startup tries to circumvent US restrictions to access Nvidia chips
- Claude can now access bank accounts to analyze spending
Google changes AI rules for schools
The lawsuit comes at a time of growing pressure over the use of AI in educational institutions.
This week, Google announced changes to the settings of its tools for school accounts. The company intends to disable Gemini by default for students, allowing educational institutions to decide whether they want to maintain access.
The measure also involves AI features embedded in Search, such as automatically generated summaries. According to Reuters, schools that wish to preserve students’ access to the search function will still have to contend with these features.
The debate is not limited to the teenage girl’s case. A survey by Common Sense Media had already pointed out risks related to the use of AI tools by children and teenagers.Among the concerns are the difficulty of systems in identifying indirect signs of suicidal thoughts and the possibility of responses that normalize behaviors associated with eating disorders.
Specialists also warn that prolonged conversations with chatbots can create a false sense of support. As a result, children and adolescents may believe they are receiving enough support and delay seeking help from family members or specialized professionals.
Chatbot safety returns to the center of the debate
The lawsuit against Google adds to other legal actions that question technology companies’ responsibility for the effects of their chatbots.
In 2025, the parents of teenager Adam Raine sued OpenAI, alleging that ChatGPT contributed to the death of their 16-year-old son by providing information related to suicide methods.
Character.AI also faced lawsuits after the death of 14-year-old Sewell Setzer III, who developed an emotional relationship with an AI-controlled character.
In January 2026, Google and Character.AI agreed in principle to end several lawsuits related to these cases, although this did not necessarily constitute an admission of liability.
The new allegations reinforce the debate over the limits of AI in environments frequented by minors, especially when the tools begin to be used as confidants on emotional and psychological matters.
Rodrigo Mozelli is a journalist graduated from São Paulo Methodist University (UMESP) and currently works as a writer at Olhar Digital.
