Teenager in Quilmes, Argentina, Reportedly Consulted ChatGPT While Planning School Shooting Against Classmates

Argentine authorities reported that a 15-year-old in Quilmes allegedly discussed a planned school shooting with ChatGPT, including tactical preparations, before an FBI alert prompted an investigation by the Argentine Federal Police. Police later seized phones and tactical clothing; no firearms or ammunition were reported. Public reporting does not establish what ChatGPT answered or whether it materially assisted the alleged plan.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.2 Cyberattacks, weapon development or use, and mass harm

The user attempted to leverage the AI system to plan, acquire tactical gear for, and execute a mass casualty school shooting.

Additional risk subdomains

• 1.2 Exposure to toxic content: The AI system was queried for and potentially exposed the user to advice or planning assistance regarding illegal acts and mass violence.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The risk and planned harm were driven by the intentional actions of a human user leveraging a deployed AI system to plan a violent crime.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The system involved is ChatGPT, which is a general-purpose conversational chatbot subject to transparency and safety monitoring obligations.

AI system and alleged parties

• AI system: ChatGPT (OpenAI)
• AI purpose: Chatbot; Writing Assistant
• Behaviour type: Assistant
• Alleged developer: OpenAI, Large language model developers, Chatbot developers
• Alleged deployer: Unnamed adolescent boy in Quilmes, Argentina, Students, School violence planners and assailants, Minors, Men and boys, ChatGPT users, Chatbot users, Adolescents, Adolescent boys
• Alleged harmed parties: Victims of school violence planners and assailants, Teachers, Students, Minors, Educators, Educational communities, Classmates of unnamed adolescent boy in Quilmes, Argentina

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Substantial. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Substantial, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: The report describes that the user generated queries and held conversations with the AI regarding how to execute a school shooting.

Directly caused: The user entered prompts detailing plans to murder classmates and requested tactical advice.

People affected

• Occurrences reported: 1
• People reportedly exposed: 30

Potential causes

Management

• Reliance on Human Moderation Loop: OpenAI's safety pipeline required manual review before law enforcement escalation.

Technology

• AI Advisor Availability: ChatGPT was accessible and used by the minor to plan a school shooting.
• Failure to Block Harmful Queries: The AI model processed queries on how to carry out a school massacre.

Data Inputs

• User Prompting for Violence: The teenager entered specific prompts detailing intent to murder classmates.

Human Factors

• Teenager's Violent Intent: The 15-year-old harbored a desire to carry out a school shooting.
• Acquisition of Tactical Gear: The minor actively acquired tactical clothing and equipment for the attack.

Process and Methods

• Inadequate Real-Time AI Blocking: The platform relied on post-hoc detection rather than blocking the generation.
• Dependence on External Alerts: Local intervention relied entirely on OpenAI and FBI detection channels.

Regulatory Environment

• Legal Inimputability of Minors: The 15-year-old is legally exempt from prosecution under local juvenile laws.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts of the incident from official police and judicial sources in Argentina, including the specific dates, location, items seized, and the mechanism of the AI safety trigger and subsequent law enforcement intervention.

A 15-year-old in Argentina used ChatGPT to plan a school shooting. OpenAI safety systems flagged the activity, leading to an FBI referral and a successful raid by Argentine police. The incident represents a localized public safety threat successfully managed by standard law enforcement and automated safety protocols with minor national security implications.

• Overall national security impact: Minor
• Scope: Multiple nations
• Primary target: Argentina
• Other affected: United States
• Alleged perpetrator: Individual

Threat characteristics

• Imminence: Long-term. The planned attack was scheduled for 2027, representing a long-term threat that was successfully neutralized years in advance.
• Autonomy: Human-controlled. The AI acted solely as a conversational assistant to a human user who made all decisions and attempted to acquire physical gear.
• Novelty: Evolved capability. Represents an evolved use of public LLMs by malicious actors to assist in planning physical violence, triggering cross-border law enforcement intervention.

Impact by dimension

• Physical security: Minor. The incident involved planning a localized school shooting rather than a threat to critical infrastructure or national-scale physical security. While grave, local police managed the threat.
• Information security: Minor. The incident triggered international intelligence sharing between OpenAI, the FBI, and Argentine federal police, but did not compromise classified information or involve systematic information warfare.
• Sovereignty: Negligible. No threat to state authority, electoral systems, or core government decision-making processes was observed.
• Economic security: Negligible. There was no impact on strategic industries, financial systems, or national technological competitive advantage.
• Societal stability: Minor. The planned attack targeted a local school community, creating localized distress, but did not threaten large-scale societal stability or cause population-scale disruption.