A cybercriminal used AI agents to commit a string of attacks against more than 100 different companies — an alarming incident that shows how autonomous agents with advanced AI hacking capabilities have exponentially raised the threat that a single human hacker can pose.
As Reuters first reported, a threat researcher at the cybersecurity company NorForge discovered that a Chinese-speaking hacker had used autonomous agents to carry out the mass hacking spree, which impacted global companies including John Deere, SharkNinja, and Cummins, among dozens of others.
The hacker isn’t alone. According to Snorre Fagerland, the researcher who discovered the Deere and SharkNinja hacker, the cybercriminal is one of several malicious actors who have abused an open source AI tool called Cairn to send swarms of agents on autonomous hacking sprees.
“It’s a new world now,” Fagerland, who detailed his findings in a report, told Reuters.
One of the Cairn-assisted hackers who Fagerland observed was able to make away with sensitive credit card information, according to the report.
The Chinese speaking hacker was the most aggressive, per Reuters. They deployed AI systems against over 1,200 US, Canadian, Australian and Japanese e-commerce sites, with successful breaches occurring after the AI agents attacked different companies for about 23 hours. Each successful hack reportedly cost about $50 in tokens, and according to Fagerland, the techniques used by the agents were sophisticated.
“It’s not straightforward to come up with solutions like that,” Fagerland told Reuters. “It’s not script kiddie stuff.”
Deere told Reuters that it had taken “necessary remediation actions” in response to the breach, and said that there was “no indication that sensitive information was accessed.” SharkNinja told the outlet that that the hacker had accessed a site that didn’t contain sensitive data, while Cummins relayed that it was “investigating the report of a potential cyber-attack on a website operated by one of our third-party vendors.”
The report is the latest to shed light on the global cybersecurity crisis wrought by advanced AI and its ability to find and exploit existing vulnerabilities in the code maintaining the world’s digital infrastructure. If the web is made of sticks and glue, AI has handed bad actors bundles of cheap TNT to plant across the internet. And while AI agents are helping normies send emails, book dinner reservations, and manage calendars, they’re also helping cybercriminals set autonomous swarms of capable, always-on hackers loose on online world.
More on AI and cybersecurity: Frontier OpenAI Model Exposes Stunning Exploit That Leaves Millions of Web Servers Open to Attack