Anthropic report: Claude targeted by Russia linked campaign, China labs
Anthropic, a major player in AI safety, just revealed that its Claude models were targeted by a suspected Russia-linked cyber espionage campaign and by China-based AI firms accused of trying to extract and replicate Claude's capabilities over the past eight months.
According to its latest Threat Intelligence report, cybercriminals and state-backed hackers are now using AI to orchestrate and execute large portions of cyberattacks.
Chinese labs accused of extracting Claude
Seven Chinese labs, including Alibaba, Moonshot, DeepSeek, and Xiaomi, are accused of tapping into Claude's features to boost their own AI projects.
Alibaba reportedly ran a campaign with more than 151 million exchanges to extract tech for its Qwen models.
Meanwhile, Moonshot and DeepSeek allegedly routed live customer conversations, which sometimes included sensitive information, through Claude and used its responses as training data.
On the Russian side, a hacking group whose tradecraft was consistent with Russia-based threat actor Midnight Blizzard used AI at nearly every stage in phishing, hotel Wi-Fi hijacking and WhatsApp-takeover operations against Ukrainian government, military and diplomatic targets.
Anthropic says it disrupted attacks
Anthropic says it had disrupted what it described as attacks over the past eight months.