AI Agents Attempted Attacks on 1,200 Websites

A Chinese-speaking operator used artificial intelligence agents to attempt attacks on more than 1,200 websites, with more than 100 reported breaches, cybersecurity researcher Snorre Fagerland said. The campaign, which began September 25, targeted online businesses in the United States, Canada, Australia and Japan, including sites associated with Deere, SharkNinja and Cummins. Fagerland, a principal threat researcher at NorForge, observed the activity through exposed instances of Cairn, an open-source platform that coordinates AI agents to solve problems. His report said the operator gave the agents brief instructions outlining intrusion tactics, then allowed them to probe targets with limited human involvement. The campaign’s reported counts and operational details are based chiefly on Fagerland’s observations. The agents attempted to find ways to gain access to websites and, in successful cases, typically spent about 23 hours probing for weaknesses, according to Fagerland. In one example he described, the agents failed to find a vulnerability in a company’s public-facing application, searched digital records for related infrastructure, then located an exposed control panel...