A hacker broke into John Deere and SharkNinja with an army of AI agents

theverge.com
October 19, 2026
2 min read
44/100
Summary
Norforge, an AI security firm, found exposed work from Cairn, an open-source AI platform, indicating that a hacker used largely autonomous AI agents to breach more than 100 websites. Reuters reported that the affected organizations included John Deere, SharkNinja, and Cummins. Norforge identified the attacker as the most active operator among exposed Cairn instances and said the campaign targeted companies in the United States, Canada, Australia, and Japan. Cairn describes itself as a general-purpose problem-solving engine that lets users define starting and ending points for AI agents, which record their progress while completing tasks. Norforge said the same system can be used to orchestrate cyberattacks. Reuters reported that the attacker specified preferred intrusion tactics in fewer than 100 Chinese characters and that successful breaches took about 23 hours. Norforge found more than 20 exposed Cairn instances it considers likely malicious and said it is working to have them taken down. John Deere told Reuters it had no indication that sensitive information was accessed. SharkNinja said a standalone product-support site was affected and that it contained the threat. Cummins said it was investigating a reported potential attack on a website run by a third-party vendor. Cairn creator @le1xia0 said the platform was never intended as a hacking tool and that misuse is a human decision.
Key Takeaways
- Norforge found evidence that a hacker used AI agents through the open-source Cairn platform to breach more than 100 websites with little human involvement.
- Reuters identified John Deere, SharkNinja, and Cummins among the organizations affected by the campaign, which targeted companies in four countries.
- Norforge found more than 20 exposed Cairn instances that it considers likely malicious and is seeking to have them taken down.
- John Deere said it had no indication sensitive information was accessed, while SharkNinja said it contained an incident affecting a standalone product-support website.
What the discussion said
The thread barely engaged with the article’s AI-agent angle. Rather than debating whether autonomous agents materially changed the scale, speed, or sophistication of the intrusion, commenters focused on identifying the underlying reporting and inferring the likely target. One reader read the cited incident report as an ordinary campaign against Magento-based e-commerce sites, aimed at stealing money rather than pursuing more exotic manufacturer-specific assets. That framing quietly deflates the headline’s dramatic image of an AI-agent army: the available discussion treats the operation as financially motivated cybercrime, not evidence of some novel AI-enabled strategic breach. There was also a hostile aside about one named victim, but it offered no argument about AI systems, agent capabilities, security implications, or the ethics of automated intrusion. Others simply supplied archived and original-source links. As a result, the conversation provides almost no basis for judging whether AI agents were genuinely central to the attack, whether they made the operation more effective, or whether this represents a broader security shift. The closest thing to consensus is that the reported target appears to have been e-commerce infrastructure and theft, while the article’s AI framing went largely unexamined.
Coverage & discussions
Hacker News · theverge.com
Community discussions
Related Articles

Anthropic AI created fake profiles and impersonated people in attempted hack
Aug 5, 2026

VMs won't contain cyber-capable agents
Aug 26, 2026

Ruby 4.0 Universal RCE Deserialization Gadget Chain
Aug 14, 2026

GPT 5.6 Cyber
Aug 10, 2026

AI assistant hacks gym website in first known Australian autonomous cyber attack
Aug 9, 2026