AI agents targeted more than 1,200 websites in hacking campaign

A Chinese-speaking hacker used artificial intelligence agents to breach websites belonging to more than 100 companies, while the agents attempted attacks on more than 1,200 sites, cybersecurity researcher Snorre Fagerland said. The campaign targeted online businesses in the United States, Canada, Japan and Australia, and included sites associated with Deere, SharkNinja and Cummins. Fagerland, a principal threat researcher at NorForge, said he observed the activity through exposed instances of Cairn, an open-source AI problem-solving platform. The hacker assigned the agents broad goals and intrusion tactics in fewer than 100 Chinese characters, then left them to probe targets with little human involvement, according to Fagerland’s report. He began finding exposed Cairn instances around October 2; the campaign itself had started around September 25. Fagerland said the agents typically spent about 23 hours probing for weaknesses during successful intrusions. In one case, they failed to find a vulnerability in a company’s public-facing application, then searched digital records for related infrastructure, located an exposed control panel and exploited a weakness to make a server...