Authorities in central Florida have arrested a 13-year-old student who typed a violent query into ChatGPT during class on a school-issued device, an incident swiftly flagged by campus monitoring software and relayed to law enforcement in a case officials said was “another ‘joke’ that created an emergency on campus.” The Volusia Sheriff’s Office said a school resource deputy at Southwestern Middle School in DeLand received an automated alert on September 26 that the student had entered the phrase “How to kill my friend in the middle of class” into the chatbot. Deputies responded to the school, questioned the boy and took him into custody. Investigators said the student told officers he had been “trolling” a classmate who had annoyed him; officials said they did not treat the matter as a prank and urged parents to speak with their children about the consequences of such actions. The sheriff’s office did not release the child’s name, and it was not immediately clear what charges he might face.
The alert that set the response in motion came through Gaggle, a safety platform used by K-12 districts to monitor student activity on school-managed accounts and devices for content that may indicate self-harm, threats or other risks. According to the sheriff’s account reviewed by reporters, the system surfaced the exact wording of the boy’s ChatGPT prompt to the campus deputy, who relayed the information to colleagues and initiated the on-site intervention. The sheriff’s office amplified its message with a direct appeal: “Parents, please talk to your kids so they don’t make the same mistake,” a line that followed a pattern of recent public advisories from the agency when school safety systems trigger emergency responses to student posts or searches.
Local and national outlets that examined the case reported the arrest unfolded during the school day without any injuries or broader disruption on campus. The boy was transported for processing under juvenile procedures after the initial interview, and officials said they would not release identifying details because of his age. A spokesperson for the district did not immediately respond to questions about whether the school day was interrupted or whether administrators imposed additional discipline under student conduct codes separate from any criminal process. OpenAI, which operates ChatGPT, did not immediately comment on the matter.
The core facts have been consistent across accounts from the sheriff’s office and subsequent coverage: a middle-school student used a district device to ask a violent, specific question of the chatbot; a school safety platform that scans student accounts flagged the language in real time; a school resource officer received the alert and moved to identify and detain the student; and the boy later told deputies he had meant the post as a joke directed at a friend. Investigators said the description of the exchange did not alter the response, citing the seriousness with which authorities treat any message that evokes violence in a classroom setting.