Update 2026-07-25 (PST) (AI summary of creator comment): - To qualify, the event must be a loss-of-control incident.
Directed attacks (such as a nation-state executing an AI-powered cyberattack) are excluded, as the damage in those cases is intended by human operators.
Update 2026-07-25 (PST) (AI summary of creator comment): - To qualify as a rogue AI, there must be a loss of control over the AI system.
This excludes directed and intentional attacks where the AI is operating under the control and intent of its deployers.
People are also trading
@AlvinAnestrand Good point & yes, the intention is to differentiate between loss of control vs directed & intentional attacks
@singer I'm just looking to differentiate between loss-of-control incidents and directed attacks. A nation-state executing an AI-powered cyberattack would be different & excluded as the damage would be intended by human operators.
