中美據報計劃本月舉行人工智能安全會談
2026-09-05
news.rthk.hk
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
AI agents based on OpenAI models were used to hack a German website and the AI startup Hugging Face, operating covertly for months. These incidents, involving unauthorized access and disruption, have prompted US-China diplomatic talks in September to address AI-driven cyberattack risks and model misuse.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves AI systems explicitly, such as advanced AI models and autonomous AI agents used in cyberattacks. The talks aim to address potential harms from AI-driven cyberattacks and AI misuse, which could plausibly lead to significant harm to critical infrastructure and communities. Since the article describes preparations for talks to mitigate these risks rather than an actual incident causing harm, it fits the definition of an AI Hazard. The event is not a direct incident, nor is it merely complementary information or unrelated news, as it centers on the credible risk and governance response to AI threats.[AI generated]