One hacker, AI agents and a few thousand dollars: the cyber spree that breached scores of sites

A Chinese-speaking hacker has broken into more than 100 company websites — and did almost none of the work. Snorre Fagerland, principal threat researcher at cybersecurity firm NorForge, said the cybercriminal outsourced the spree almost entirely to AI agents, which spent hours probing and breaking into targeted sites (Reuters).

From 25 September, the hacker set the agents loose against e-commerce websites in the US, Canada, Australia and Japan, with terse instructions of fewer than 100 Chinese characters. By late last week the agents had attempted to break into more than 1,200 sites, successful intrusions typically taking around 23 hours. Fagerland estimated the campaign — apparently using a mixture of China’s DeepSeek and Anthropic’s Claude Code — had consumed roughly $5,000 worth of tokens, about $50 per successful hack.

In one case the agents failed to crack a big company’s application, so they traced its associated infrastructure, found a control panel with a subtle flaw in how it processed data, then tricked the server into executing malicious commands. “It’s not straightforward to come up with solutions like that,” Fagerland said. “It’s not script kiddie stuff.”

Reuters independently identified the victims from the hacker’s Cairn activity as including Deere, SharkNinja and Cummins. Deere saw “no indication that sensitive information was accessed”; SharkNinja contained the breach of a support-info site; Cummins said it was investigating. Cairn’s author, handle “le1xia0”, did not answer Reuters’ questions, saying on X the platform was “an innovative type of multi-agent architecture and was not designed for hacking attacks”.