The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
A 16-year-old in Sydney was radicalized online after TikTok's algorithm exposed him to extremist content. He used Telegram for communication and ChatGPT to strategize a terror attack, including bomb-making research. The teen pleaded guilty but avoided jail, highlighting AI's role in facilitating terrorism planning.[AI generated]
Why's our monitor labelling this an incident or hazard?
The AI systems (TikTok's recommendation algorithm and Telegram's encrypted platform) played a role in exposing the teenager to extremist content and facilitating communication about terrorism. Although no actual terrorist attack or physical harm occurred, the AI systems indirectly contributed to the planning of a terror attack, which represents a plausible risk of serious harm. The event does not describe a realized harm but a credible potential for harm due to AI-enabled radicalization and planning. Therefore, it fits the definition of an AI Hazard rather than an AI Incident. The sentencing and deradicalization efforts are responses to the hazard but do not change the classification.[AI generated]