🔍 Read the full analysis: AI Agents Change The Picture For One Hacker Targeting 100+ Sites on ThorstenMeyerAI.com
Get smart everyday buys delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
A security researcher and Reuters documented an operation in which a Chinese-speaking operator used Cairn, an open-source AI-agent platform, to probe online shops and gain access to more than 100 company websites, according to the attacker’s records. The scale and cost figures are estimates, and it remains unclear what data, if any, was taken from most affected sites.
A Chinese-speaking operator used Cairn AI agents to probe online shops from late September through mid-October, with records reviewed by Reuters indicating successful break-ins at websites of more than 100 companies. The affected businesses included Deere, SharkNinja and Cummins; the extent of any data theft remains unclear.
According to Reuters, the agents tried more than 1,200 sites across the United States, Canada, Australia and Japan. The count of successful intrusions comes from the operator’s own Cairn records, which Reuters said it independently reviewed to identify affected companies. NorForge, the security firm whose researcher monitored exposed control panels, reported that its largest activity cluster had more than 1,100 attack projects across 14 panels, mainly aimed at the United States, Japan and Australia.
NorForge researcher Snorre Fagerland found internet-accessible Cairn dashboards, some without passwords, and observed activity through them. The firm identified more than 20 panels it linked to likely malicious use and grouped the activity into three clusters. NorForge’s report, published October 19, did not name victims; Reuters said it identified companies by examining the attacker’s Cairn activity and contacting them. Two smaller businesses acknowledged compromises, while some other companies did not respond.
The companies described differing effects. Deere told Reuters it took “necessary remediation actions” and had no indication sensitive information was accessed. SharkNinja said a standalone product-support site was affected and the threat was contained. Cummins said it was investigating a potential attack on a website run by a third-party vendor. Reuters could not establish how much information, if any, was taken.
One operator, a swarm of agents
The campaign in numbers
What the researcher and Reuters report about the busiest group NorForge found. Counts come from the attackers’ own exposed panels; cost and success figures are estimates.
The price tag estimates
NorForge / Fagerland, “uncertainty considerable”
- ~$5 per task on average, including tasks that fail early
- $100+ for long projects with more than 1,000 intents and facts
- ~$5,000 in AI costs for the campaign, about $50 per successful hack
- ~10% success rate, NorForge’s “reasonable assumption”
How the swarm runs
Simplified, no attack details
- Shop websites are ranked by traffic and checked automatically in a browser
- Each target becomes a Cairn project: a starting point and a goal
- Agents post facts and next steps on a shared board and build on each other
- Goal: a hidden way back in to shop and payment systems, for later
- 22 SepGambit reportGambitCairn found among three AI harnesses on a criminal’s staging server (separate campaign)
- 25 SepCampaign startsReutersA Chinese-speaking operator turns agents on shops in the US, Canada, Australia and Japan
- ≈ 2 OctOpen panels foundReutersFagerland starts finding Cairn panels in malicious use, about two dozen
- mid-Oct1,200+ sites triedReutersBreaks into websites of 100+ companies, incl. Deere, SharkNinja, Cummins
- 19 OctReport and investigationNorForge · ReutersNorForge publishes “Inside Cairn”; Reuters publishes its investigation
Sources: NorForge, Inside Cairn (19 Oct 2026); Reuters (19 Oct 2026); Gambit Security (22 Sep 2026). Numbers as published, same rounding as the article.
Powered by Thorsten Meyer AI
Automation Expands the Attack Window
The reports point to a change in the effort required to sustain cyber intrusions: an operator can set a goal and let multiple agents carry out repeated probing with limited direct supervision. NorForge measured successful projects in the largest cluster as taking a little under 23 hours from start to goal on average, with about 140 recorded facts per project. Those measurements describe the observed activity, not a general benchmark for AI-enabled attacks.
NorForge estimated the campaign’s AI costs at roughly $5,000 overall, or about $50 per successful intrusion, while warning that the estimate has considerable uncertainty. If accurate, low operating costs and the ability to keep agents working could make broad, persistent targeting more practical. The report’s interpretation was that the operator sought persistent access to e-commerce or payment infrastructure for possible later monetization; that goal is an assessment, not proof of what happened to data or access after the intrusions.
For organizations, the reported targets included not only major company websites but also a product-support site and a third-party vendor’s site. That makes visibility into smaller online properties and vendor systems relevant to incident response. The reporting does not establish that AI agents independently made every decision: an operator supplied instructions, and Cairn coordinated the agents’ work.
Popular right now
Ninja Hot & Iced XL Coffee Maker with Rapid Cold Brew, Black, CM371
Ninja DualZone Air Fryer 10 Qt Smart 6-in-1 Grey (DZ950)
Vitamix 5200 Professional-Grade Blender, 64 oz. Container, Black
As an affiliate, we earn on qualifying purchases.
How Cairn Panels Revealed Activity
The reporting followed a September investigation by Gambit Security into a separate campaign. Gambit said it recovered a criminal staging server containing three “AI harnesses”—software that lets AI models take actions—including Cairn, Hermes and Strix. NorForge then searched for exposed dashboards associated with Cairn. Its report says most panels require authorization, but some were publicly accessible.
Cairn is an open-source agent platform, not a tool designed exclusively for cyberattacks. NorForge describes it as a state-based problem-solving system: agents record what they have learned and what they plan to do, then use those notes to guide further steps. The platform’s author, who uses the name le1xia0, told Reuters that Cairn was not designed for hacking and that harmful use was not their intention.
The September Gambit report concerns a separate operation, not the campaign Reuters and NorForge examined. Gambit reported at least 27 companies compromised and at least 600,000 card records taken from two of them. That earlier case offers context for agent-driven criminal activity but does not establish that the same operator or methods were involved in the later campaign.
““Necessary remediation actions””
— Deere, in a statement to Reuters
Data Loss and Operator Identity
The operator’s identity has not been established. NorForge characterized the activity as linked to a Chinese-speaking actor; that description does not identify a person, group or location. The supplied reporting states that there was no reported arrest or charge.
It is also unclear whether information was taken from most affected websites, whether attackers retained access, or whether any access was sold or used for fraud. Reuters could not determine the amount of data involved. Deere reported no indication that sensitive information was accessed, but that statement does not settle what occurred at other companies. The AI-cost estimate and per-intrusion figure are NorForge’s estimates, and the firm cautioned that the uncertainty is considerable.
NorForge’s project and panel counts, Reuters’ tally of company websites, and company statements measure different parts of the activity. They should not be treated as interchangeable counts of confirmed data theft or confirmed long-term access.
Companies Continue Incident Reviews
The next developments depend on affected organizations’ investigations and any further disclosures about access or data exposure. Cummins said it was investigating a potential attack involving a third-party vendor’s website. Deere and SharkNinja described response steps, but the available reporting does not provide a shared account of the full scope or duration of each incident.
NorForge’s report identified exposed Cairn control panels and several activity clusters, while Reuters’ investigation connected one cluster to named companies. Further technical analysis or company updates could clarify how access was gained, whether it persisted, and whether the same operator targeted additional sites. No arrest, charge or confirmed account of stolen data was reported in the source material.
Key Questions
What happened in the reported campaign?
A Chinese-speaking operator used Cairn AI agents to probe online shops. Reuters reported that the agents tried more than 1,200 sites and that the operator’s records showed break-ins at websites of over 100 companies.
Which companies were named?
Reuters named Deere, SharkNinja and Cummins among the companies whose websites were affected. The reports also said two smaller businesses acknowledged compromises.
Was customer or company data stolen?
The amount of data taken, if any, was not established. Reuters could not determine how much information was accessed. Deere said it had no indication sensitive information was accessed; that statement does not address every affected site.
Who was behind the operation?
The operator’s identity is unknown. NorForge described a Chinese-speaking actor, but the reporting did not identify a specific person or group, and no arrest or charge was reported.
What is Cairn?
Cairn is a free, open-source platform that coordinates AI agents through a shared record of findings and planned steps. Its author said it was not designed for hacking; the reports describe its alleged use in this campaign.
Source: ThorstenMeyerAI.com
