CyberHighReportedBuilding
6.6
Anthropic warns Chinese open-weight model GLM-5.3 poses elite hacking risk
South China Morning PostLO·US · CN·1 day ago
Recent weeks have seen rising concerns about the trustworthiness of AI, particularly regarding its potential use in cyber attacks. The item discusses general worries rather than a specific incident, leaving the actual threat level uncertain. This matters as AI-driven cyber threats could escalate risks to critical infrastructure and corporate systems.