Researchers Watch AI Agents Break Into Over 1,200 Company Websites Through Exposed Control Panels

Security researcher Snorre Fagerland of NorForge discovered a criminal operation using autonomous AI agents to attack company websites after finding the attackers' control panels left open on the internet.

The agents attempted entry at more than 1,200 sites and reached websites belonging to over 100 companies, including Deere, SharkNinja and Cummins, according to Reuters, which reviewed the attacker's records. It is not known how much data, if any, was taken.

Fagerland found about two dozen exposed instances of Cairn, an open-source AI tool, beginning around October 2, letting him observe the attackers' work in real time. Cairn gives AI agents a starting point and a goal, then lets them work step by step, posting findings to a shared board where a human operator can add hints. Its author, who posts as "le1xia0," told Reuters the tool was not designed for hacking and that harmful use was not intended.

NorForge grouped the panels into three clusters and judged more than 20 likely malicious. The busiest, called AIA0001, is Chinese-language and links 14 panels to more than 1,100 attack projects, apparently aimed at lasting access to shop and payment systems, with targets in the United States, Japan and Australia. NorForge confirmed one claimed break-in at a large hardware-parts vendor and independently verified the compromise of one very large US retailer.

Deere said it fixed the problem and saw no sign sensitive information was accessed. SharkNinja said the attack hit a standalone product-support site and was contained. Cummins said it was investigating a report of a possible attack on a website run by one of its vendors. Two smaller businesses also told Reuters they were compromised.

Break-ins took just under 23 hours on average, with a success rate around 10%, NorForge found. It estimated the average cost at about $5 a task, though it cautioned the figure is rough; Fagerland told Reuters the whole campaign used about $5,000 worth of tokens, or roughly $50 per successful hack.

The open panels expose victims' secrets as well. NorForge ran the Trufflehog scanner over panel data and found hundreds of secrets, 11% of which it verified as still valid. It also found apparently legitimate penetration-testing firms that had left real client tests on unsecured panels. NorForge said it is working with Reuters to warn victims and asking hosting firms to take the panels down.

Source: The Hackers News