🔍 Read the full analysis: How AI Agents Alter The Economics Of Website Hacking 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 Chinese-speaking operator used Cairn, an open-source platform for coordinating AI agents, in a campaign targeting online shops from late September to mid-October 2026. Reuters counted breaches at websites of more than 100 companies from the attackers’ records; NorForge estimated the campaign cost about $5,000, but the data taken and the operator’s identity are unknown.
A Chinese-speaking operator used AI agents to probe online shops from September 25 through mid-October 2026, with Reuters counting breaches at websites of more than 100 companies from the attackers’ own records. The campaign used Cairn, a free, open-source agent-coordination platform, and shows how automation can let one operator run prolonged attempts against many sites; how much data was accessed or taken remains unknown.
Security researcher Snorre Fagerland of NorForge found internet-facing Cairn control panels and tracked activity through some panels that lacked password protection. NorForge linked more than 20 exposed panels to likely malicious use and grouped activity into three clusters. Its largest, called AIA0001, used Chinese-language instructions. NorForge counted more than 1,100 attack projects across 14 panels, targeting sites mainly in the United States, Japan and Australia.
Reuters separately reviewed Cairn activity and identified affected companies, including Deere, SharkNinja and Cummins. Deere told Reuters 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 said it could not establish how much data, if any, had been taken.
NorForge measured successful AIA0001 projects as taking a little under 23 hours from start to goal, with one project lasting more than five days. The firm estimated the campaign’s AI costs at roughly $5,000, or about $50 per successful hack, while warning that the estimate carried considerable uncertainty. Reuters’ figure of more than 100 affected companies is based on the attackers’ records, not a complete public accounting of confirmed victims.
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
Why Automated Patience Matters
The reporting suggests that AI agents can lower the labor required to keep testing targets over extended periods. NorForge’s findings describe software sharing information about what it has found and what it intends to try next, allowing the work to proceed across multiple agents rather than requiring an operator to direct every step. The reported cost estimate is not a verified bill, but it points to a potential economic shift: attackers may be able to spend more time probing each target for comparatively little money.
For companies, exposure may extend beyond their main storefront. Reuters’ reporting includes a product-support site and a website run by a third-party vendor, illustrating why less prominent services can carry security risks. The campaign also creates a workload problem for defenders: automated activity can generate many findings that must be checked and prioritized. NorForge’s account raises a further question about how organizations can identify and control software agents that interact with their systems.
Those implications should not be mistaken for proof that AI caused every successful intrusion or that the campaign produced a specific financial return. The evidence establishes reported activity and breaches, while the extent of any data theft and the attackers’ eventual gains remain unresolved.
Products Worth Considering
Samsung 27-Inch Odyssey QD-OLED G8 4K 240Hz Gaming Monitor
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
How Cairn Activity Was Tracked
The investigation followed a separate report by Gambit Security in September 2026. Gambit said it recovered a criminal staging server containing three AI “harnesses”—programs designed to let models take actions—including Cairn. NorForge then searched for exposed Cairn dashboards, which can serve as control panels for coordinated agents. Fagerland found panels that were accessible without a password, although NorForge said most required authorization.
NorForge published its report on October 19, 2026, the same day Reuters published its investigation. NorForge’s report describes activity and clusters but names no victims; Reuters says it independently reviewed the attacker’s Cairn activity and contacted companies. Two smaller businesses acknowledged compromises, Reuters reported. The distinction matters: the firms’ technical analysis and the independently identified victim list come from different reporting steps.
Cairn is not described by NorForge as a purpose-built hacking product. Its author, who uses the name le1xia0, said the platform was not designed for hacking attacks and that harmful use was not the author’s intention, according to Reuters. In a separate September report, Gambit documented another agent-driven campaign, reporting at least 27 companies compromised and at least 600,000 card records taken from two of them. That was a different campaign and should not be treated as evidence of data theft in the Cairn activity Reuters and NorForge examined.
““[Cairn] was not designed for hacking attacks. If it is used to carry out harmful hacking attacks, that is not my intention.””
— Cairn author le1xia0, quoted by Reuters
What Victim Impact Remains Unknown
The operator’s identity has not been established publicly. The source material describes the person behind the activity as a Chinese-speaking actor; it does not establish who that person is, their location, or whether they acted for a group or organization. No arrest or criminal charge is reported.
The available accounts also do not establish how much data, if any, was taken, whether stolen information was later used or sold, or what financial return the operator achieved. Reuters could not determine the amount of data accessed. NorForge’s cost figure is an estimate, and the firm described its uncertainty as considerable. The reported count of more than 100 companies comes from attacker records reviewed by Reuters; the public information does not provide a complete list or independently confirm every entry.
Incident Reviews and Exposure Checks
Companies named in the reporting and their vendors will need to continue reviewing affected sites, determining whether data was accessed and documenting remediation. Cummins said it was investigating a potential attack; Reuters reported that Deere had taken remediation steps and SharkNinja said the threat was contained. The source material does not give a timetable for final findings or describe a completed law-enforcement investigation.
For security teams, the findings point to practical follow-up: review internet-exposed control panels, check protections on support and vendor-operated websites, and examine how automated agents are authorized to interact with systems. NorForge’s discovery of panels without passwords makes access controls a specific concern, although the reporting does not establish that every exposed panel was used in this campaign. Further reporting or company disclosures may clarify the number of confirmed victims and whether any information was taken.
Key Questions
What happened in the AI-agent campaign?
A Chinese-speaking operator used agents coordinated through Cairn to probe online shops from September 25 into mid-October 2026. Reuters counted breaches at websites of more than 100 companies using records from the attacker’s Cairn activity.
Which companies were identified?
Reuters identified websites associated with Deere, SharkNinja and Cummins, among others. The companies’ statements described remediation, a contained threat at a standalone support site, and an investigation involving a vendor-operated site. The public reporting does not provide a complete, independently confirmed victim list.
Was customer or company data stolen?
That has not been established. Reuters could not determine how much data, if any, was taken. Deere said it had no indication that sensitive information was accessed; that statement does not settle the impact at other affected sites.
How much did the campaign cost?
NorForge estimated roughly $5,000 in AI costs, or about $50 per successful hack. The firm described the estimate as highly uncertain, so it should be treated as an estimate rather than a verified campaign expense.
Was Cairn built for hacking?
NorForge describes Cairn as a general-purpose, state-based problem-solving engine. Its author, le1xia0, told Reuters it was not designed for hacking attacks and said harmful use was not the author’s intention.
Source: ThorstenMeyerAI.com
