Anthropic says Claude was used in cyberattacks and surveillance
Claude AI cyberattack report exposes how operators automate hacks and research See the risks behind state-linked misuse and what Anthropic found in its case studies
Anthropic said in a recent report that its Claude AI model was used by Russian and Chinesespeaking operators to help automate cyberattacks and security research.
According to the company, one Russianspeaking actor used AIdriven workflows to speed up attacks against more than 20 organizations, including government ministries, intelligence bodies, embassies, and diplomatic missions in Ukraine and Europe. Anthropic also said Chinesespeaking operators used Claude to support vulnerability research, including work that generated possible zeroday findings in firmware.
The report also described a separate case in Mali, where a consultant reportedly used Claude to help build a large domestic surveillance system for state intelligence services. Anthropic said the case shows how AI can reduce the time and effort needed for both cyber operations and mass surveillance.