By David Jones
Publication Date: 2025-11-14 10:52:00
Anthropic said a suspected state-affiliated hacker manipulated one of its agent AI-based coding tools to carry out a sophisticated espionage campaign against about 30 major organizations around the world in September, according to a blog post published Thursday.
The hackers used the company’s Claude Code to target a range of organizations, including chemical manufacturers, major technology companies, financial institutions and government agencies. According to Anthropic, the threat actor, which the company has named GTG-1002, successfully breached a small number of the targets.
The company claims that the attack may be one of the first large-scale cyberattacks without significant human involvement. According to Anthropic, between 80 and 90% of attacks were carried out using AI, with human intervention required at four to six key decision points.
Anthropic said it suspended a number of accounts associated with the…
