A Chinese-speaking criminal used AI agents to attack over 1,200 company websites with little help. Any business with a public site should check that it is updated.
A Chinese-speaking criminal pointed autonomous AI agents at company websites and let them break in with almost no human help. The agents tried more than 1,200 sites and got into the websites of more than 100 companies. A security researcher saw it live because the attackers left their control panels open on the internet.
The researcher, Snorre Fagerland of the security firm NorForge, found the activity around October 2. He was scanning for exposed copies of Cairn, an open-source AI tool popular with hackers. It gives AI agents a starting point and a goal, then lets them work step by step. Reuters reviewed the attacker's records and named victims including Deere, SharkNinja and Cummins.
The operator's instructions were short, under 100 Chinese characters according to Reuters. NorForge found successful break-ins took a little under 23 hours on average, with a success rate around 10%. Cairn's author told Reuters the tool was not designed for attacks.
The cost was low. NorForge put the average at about $5 a task, though it called the figure rough. Fagerland told Reuters the whole campaign used about $5,000 worth of computing, or roughly $50 per successful hack. That makes it cheap to test many sites, including small ones that nobody expected to be targeted.
The busiest group of attackers seemed to want lasting access to shop and payment systems for later use. Targets were in the United States, Japan and Australia. One documented break-in used old, outdated website code, so neglected software is the kind of weak spot to worry about.
The three named companies said the damage was limited. Deere said it had no sign that sensitive information was accessed. SharkNinja said the attack hit a standalone product-support site and was contained. Cummins said it was investigating a possible attack on a site run by one of its vendors. Two smaller businesses also told Reuters they were compromised.
Even so, any business with a public website is a possible target. Small firms with unpatched site software, extra plugins or old login pages are most exposed. The steps in the alert box above cover what to ask your web provider this week.
It is not known how much data, if any, was taken. The paths in differed from victim to victim, so there is no single fix. The AI models involved were not confirmed. Reuters said only that the campaign appeared to use DeepSeek and Claude Code.
NorForge said it is working with Reuters to warn victims and is asking hosting firms to take the exposed panels down. Whether that works depends on how the hosts respond. Some of the open panels also belonged to apparently legitimate security-testing firms, which put their clients' data at risk too.
Cheap AI attacks make an out-of-date website a real risk, so get your site updated and your admin logins protected now.
A viral ChatGPT prompt turns what the chatbot remembers about you into a room image. Posting it publicly could give scammers clues to target you.
Read more: Viral ChatGPT room image could hand scammers clues about youDARPA wants AI to hunt for design flaws in stock, bond, energy and cloud markets, so enemies cannot use them to cause economic damage that reaches households.
Read more: Pentagon wants AI to find weak spots in stock and energy marketsTuring Award winner Richard Sutton says more computing power alone will not make AI truly smart. Big AI companies disagree, which matters for anyone betting on AI.
Read more: AI pioneer says piling on computing power won't make machines smart
