Anthropic disrupts Russian and Chinese AI campaigns

Anthropic disrupted 7 malicious AI campaigns targeting its Claude models over an eight-month window, including state-backed espionage originating from Russia and unauthorized data extraction by firms in China aiming to replicate its technology. The cyberattacks demonstrate how threat actors increasingly deploy multi-agent frameworks to automate operations rather than relying on manual chatbot queries. The security crackdown underscores mounting vulnerabilities as generative AI systems become central targets for intellectual property theft and automated cyberwarfare.