🔍 Read the full analysis: What Happens When One Hacker Uses AI Agents On 100+ Company Websites? on ThorstenMeyerAI.com
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A Reuters investigation, drawing on records from exposed Cairn AI-agent control panels, reported that an operator’s agents broke into websites linked to more than 100 companies after probing over 1,200 sites. Researchers estimate the campaign cost about $5,000, but the estimate is uncertain; the amount of data accessed or taken has not been established.
A Chinese-speaking operator used AI agents coordinated through Cairn to probe more than 1,200 websites and break into websites associated with more than 100 companies, Reuters reported, citing the operator’s own records and an independent review. NorForge researcher Snorre Fagerland observed exposed control panels used in the campaign; the amount of data accessed or taken remains unknown.
The activity began around September 25, according to the source material, and targeted online shops in the United States, Canada, Australia and Japan. NorForge’s report, published October 19, identified several clusters of suspected malicious use of Cairn panels. Its largest cluster, labeled AIA0001, recorded more than 1,100 attack projects across 14 panels. Reuters separately reviewed the operator’s Cairn activity and identified company websites linked to the intrusions; NorForge’s report did not name victims.
Fagerland found the panels while searching for internet-accessible Cairn dashboards. NorForge said it located more than 20 panels it linked to likely malicious activity, some without password protection. Cairn lets agents record findings and planned actions on a shared, persistent workspace. The reporting describes the operator giving instructions and the agents carrying out extended probing, rather than a person directing every step manually.
Reuters reported that affected companies included Deere, SharkNinja and Cummins. Deere said it took “necessary remediation actions” and had “no indication that 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 cyber-attack on a website operated by one of our third-party vendors.” Reuters could not establish how much data, 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 Changes the Cost of Intrusions
The report matters because AI agents can keep working across a long series of targets with less direct attention from an operator. NorForge measured successful AIA0001 projects at a little under 23 hours from start to goal, with one project running for more than five days. The firm estimated AI costs of about $5,000 for the campaign, or roughly $50 per successful intrusion, while warning that uncertainty around the estimate was considerable.
That cost figure is an estimate, not a verified accounting of the operator’s spending. Still, the reported workflow points to a challenge for defenders: automated activity can increase the volume and persistence of probing, while intrusions may involve less prominent support or third-party sites rather than a company’s main public-facing service. A compromised site can create risk even when it is not clear that sensitive information was accessed.
The reporting also raises a practical security question about exposed agent dashboards and control panels. Cairn is a general-purpose platform, not a hacking product, but its shared state and autonomous task handling can be used for harmful activity. Organizations may need to account for internet-accessible systems run by vendors as well as their own central websites.
Products Worth Considering
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1
As an affiliate, we earn on qualifying purchases.
How Researchers Traced the Campaign
The investigation followed a separate September report by Gambit Security, which described a different agent-driven campaign and found three AI harnesses, including Cairn, on a criminal staging server. Gambit said that separate campaign compromised at least 27 companies and involved the theft of at least 600,000 card records from two of them. Those findings concern a distinct operation and do not establish that card data was taken in the campaign Reuters later examined.
NorForge then searched for Cairn dashboards exposed to the internet. Fagerland observed panels that researchers associated with suspected malicious activity and grouped the activity into three clusters. Reuters’ investigation added victim identification by reviewing Cairn activity and contacting organizations. The source account says two smaller businesses acknowledged compromises, while some other victims did not respond.
Cairn’s creator, who uses the name le1xia0, said the platform was not designed for hacking. The distinction matters: the platform’s stated purpose and the uses researchers observed are not the same. The findings describe how an available tool was used, not proof that its creator intended or directed the campaign.
““Necessary remediation actions” were taken, and the company had “no indication that sensitive information was accessed.””
— Deere, in a statement to Reuters
Data Access and Attribution Remain Unclear
It is not known how much data, if any, was accessed or taken in the campaign reported by Reuters. Deere said it had no indication sensitive information was accessed, while the statements summarized for SharkNinja and Cummins did not establish whether information was removed. Reuters could not determine the overall data impact.
The operator’s identity has not been established. The source material describes the actor as Chinese-speaking; that does not confirm the person’s nationality, location or identity. Nor does the reporting establish an arrest or charge. The estimated campaign cost also carries substantial uncertainty, according to NorForge, and should not be treated as a confirmed bill or a precise measure of the cost of each intrusion.
Company Reviews and Further Findings
The companies named in the reporting may provide more information as their investigations continue, particularly about affected systems, access to data and remediation. Cummins has said it is investigating a vendor-operated website; the available reporting does not give a completion date or outcome.
NorForge’s published findings and Reuters’ review provide a record of the observed campaign, but further details could change what is known about its scope. For now, the confirmed picture is limited to the activity visible in exposed Cairn panels, the company responses reported by Reuters and the researchers’ estimates. Attribution, data impact and the full number of affected organizations remain unresolved.
Key Questions
What happened in the reported campaign?
Reuters reported that an operator used Cairn AI agents to probe more than 1,200 websites and break into websites associated with more than 100 companies. The figures were based on a review of the operator’s activity records.
Who was behind the activity?
The operator has not been identified. Researchers described the activity as Chinese-language or linked the actor to Chinese-language instructions; that does not establish the operator’s identity or nationality.
Was customer or payment data stolen?
The available reporting does not establish how much data, if any, was taken in this campaign. Reuters could not determine the data impact, and Deere said it had no indication sensitive information was accessed.
What is Cairn?
Cairn is an open-source platform for coordinating AI agents through a shared, persistent workspace. Researchers said it was used in the campaign, while its creator said it was not designed for hacking.
How reliable is the estimated $50 cost per intrusion?
It is NorForge’s estimate, based on its assessment of the campaign’s AI costs and reported successful projects. The firm said the estimate involved considerable uncertainty, so it is not a confirmed cost for each intrusion.
Source: ThorstenMeyerAI.com
