Key Takeaways
- Anthropic reports first major AI-powered cyberattack using Claude chatbot
- Meta’s AI chief Yann Lecun dismisses study as ‘dubious’ and fear-mongering
- Chinese hackers allegedly used AI for 80-90% of espionage campaign
- China denies allegations, calling them ‘groundless accusations’
Anthropic has revealed that its Claude AI chatbot was weaponized by Chinese state-sponsored hackers in what the company calls the first large-scale autonomous cyber espionage campaign. The attack targeted approximately thirty global organizations including major tech firms, financial institutions, and government agencies.
The AI startup’s disclosure sent shockwaves through the technology industry, raising concerns about the potential misuse of advanced AI systems for malicious purposes.
Meta’s Chief Scientist Questions Credibility
Meta’s Chief AI Scientist Yann Lecun, a Turing Award winner and one of the “Godfathers of Deep Learning,” has strongly criticized Anthropic’s study, labeling it as ‘dubious’ and accusing the company of fear-mongering to achieve ‘regulatory capture.’
In response to calls for increased AI regulation following the report, Lecun stated: “You’re being played by people who want regulatory capture. They are scaring everyone with dubious studies so that open source models are regulated out of existence.”
This isn’t the first time Lecun has clashed with Anthropic leadership. Months earlier, he had described Anthropic CEO Dario Amodei as an ‘AI doomer’ and accused him of being “intellectually dishonest and/or morally corrupt.”
Anthropic’s Detailed Claims
According to Anthropic’s blog post, the company first detected suspicious activity in September 2025 that evolved into a sophisticated espionage operation. The hackers leveraged Claude’s AI capabilities to automate 80-90% of the attack campaign, requiring only occasional human intervention.
“At the peak of its attack, the AI made thousands of requests, often multiple per second — an attack speed that would have been, for human hackers, simply impossible to match,” the company reported.
However, Anthropic acknowledged limitations in the AI’s performance, noting that Claude occasionally hallucinated credentials or claimed to extract secret information that was actually publicly available. “This remains an obstacle to fully autonomous cyberattacks,” the company added.
China’s Response
China’s Ministry of Foreign Affairs spokesperson Lin Jian has denied the allegations, describing them as “groundless accusations that have no evidence” during a press briefing.
