Anthropic report details misuse of Claude in cyberattacks and surveillance

Anthropic released a threat intelligence report stating that Russian and Chinese actors are using its AI model, Claude, to automate cyberattacks. Russian operators, such as 'JackPoterz,' have used AI-driven workflows to target over 20 organizations, including government agencies and embassies. Chinese users have utilized Claude for vulnerability research, identifying multiple potential zero-day vulnerabilities in network equipment firmware. Additionally, the report revealed that an independent consultant in Mali used Claude to help build a nationwide surveillance system capable of monitoring 25 million SIM cards without court orders. Anthropic noted that AI is significantly lowering the barrier to entry for cyberattacks, allowing single operators to conduct intrusions in just a few hours.