How Claude became a hacking tool for 17 organizations

Anthropic company released an analytical security report. From it becomes clear that Claude and other AI agents are becoming tools of cybercriminals. At Anthropic, they called this new direction vibe-hacking. It turns out that artificial intelligence has radically lowered barriers to entry into criminal activity.

The scale of the problem is shown through specific cases. Claude Code was used to hack 17 organizations in different countries. Government agencies, medical institutions, emergency services became victims of attacks. The AI agent simultaneously acted as technical consultant and active hack operator.

Claude also generated psychologically crafted extortion demands and assessed the value of stolen data. What is extremely difficult to do manually, artificial intelligence accomplishes in minutes.

Monetization of evil also reached industrial scales. One user created ransomware viruses through AI agents for development. Then sold them to other criminals at prices from $400 to $1200.

Claude also helped DPRK specialists fraudulently get remote jobs at Fortune 500 companies. Previously, people with higher education and English knowledge were required. Now those who can’t program and don’t understand the language get jobs!

A Telegram bot created with artificial intelligence composed convincing and complimentary messages. Non-English-speaking criminals connected with victims from USA, Japan and South Korea for subsequent extortion.

Therefore Anthropic is now developing abuse detection tools, blocking dishonest users, passing information to law enforcement. But criminals find ways to bypass protection faster than these barriers are created!