Unit 42 discloses that China-linked attackers breached 10 organizations with AI-driven attacks
China-Nexus Actors Breach 10 Organizations in AI-Driven Attacks
Palo Alto Networks' Unit 42 disclosed that China-linked attackers used the open-source AI penetration testing framework Cairn to automate attacks on at least 19 organizations, 10 of which suffered unauthorized access or data theft. Operators only need to provide a target and high-level goals, and Cairn plans reconnaissance, exploitation, and records results on its own; this weaponization used Zhipu GLM 5.2 and DeepSeek V4 Flash and Pro. The 25 operations from August to September took anywhere from 20 minutes to over 50 hours each; the attackers leaked over 1 million personal records at one organization and tampered with an in-progress industry standard voting document. Targets included 7 semiconductor and advanced manufacturing supply chain companies, 3 U.S. defense and defense policy institutions, and human rights and civil society organizations. Unit 42 said the initial intrusions exploited known vulnerabilities such as authentication bypass and weak passwords, and that traditional defenses such as identity governance, attack surface reduction, and patch management remain effective.
Unit 42's analysis shows that these AI-automated attacks still relied on known vulnerabilities, and traditional defenses such as identity governance and patch management remain effective.