🔍 Read the full analysis: A Look At AI Agents And Low-Cost Website Break-Ins on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A Chinese-speaking operator used Cairn, a free, open-source platform for coordinating AI agents, to probe online shops from late September into October, according to Reuters and security firm NorForge. Reuters reported that the attacker’s own records showed break-ins at websites of more than 100 companies; the scale of any data theft remains unclear.
An unidentified Chinese-speaking operator used AI agents to probe online shops across several countries, with records reviewed by Reuters indicating break-ins at websites of more than 100 companies. The activity began around September 25 and involved Cairn, a free, open-source platform for coordinating agents, according to Reuters and security firm NorForge; how much data, if any, was taken is not known.
By mid-October, the agents had tried more than 1,200 sites in the United States, Canada, Australia and Japan, according to the attacker’s records described by Reuters and NorForge. Reuters identified affected companies including farm-equipment maker Deere, appliance company SharkNinja and engine maker Cummins. NorForge’s report, published October 19, describes multiple clusters of exposed Cairn control panels associated with suspected malicious activity.
NorForge researcher Snorre Fagerland found internet-accessible Cairn dashboards, some without passwords, while searching for exposed agent-control panels. The firm said it identified more than 20 panels it linked to likely malicious use. Reuters separately reviewed activity in the dashboards and contacted companies; two smaller businesses acknowledged compromise, while other contacted victims did not respond.
NorForge measured successful projects in the largest activity cluster at a little under 23 hours from start to goal, on average, and reported about 140 recorded “facts” per project. Fagerland estimated the campaign’s AI costs at roughly $5,000, or about $50 per successful intrusion, while NorForge cautioned that the estimate carries “considerable” uncertainty. Those figures are estimates, not audited costs.
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 Probing Matters
The reported campaign illustrates how agent software may let one operator run many investigations in parallel, reducing the time spent manually checking sites and pursuing leads. In this case, NorForge describes an automated process that ranked shop websites by traffic, checked them in a browser and passed targets to agents. The reported average duration of just under 23 hours shows that some operations could persist and work through tasks over an extended period with limited direct supervision.
For companies, the findings point to exposure beyond the main corporate website. Reuters reported activity involving online shops, while company statements referenced a product-support site and a site operated by a third-party vendor. Sites that receive less attention than core systems can still provide attackers with a route into company infrastructure. The reports do not establish that every affected website gave access to sensitive systems or customer information.
The use of agents may also complicate monitoring: automated requests can produce many signals for security teams to review. That is an implication of the reported scale, not a measured finding about defenders’ response. The case also raises a practical question for organizations using agent platforms: whether dashboards are exposed publicly and whether access controls and monitoring cover those systems.
Products Worth Considering
FLEXISPOT Standing Desk with Monitor Riser & Cable Management
Samsung 34-Inch ViewFinity S50GC Ultra-WQHD Monitor
Dell UltraSharp 27 Thunderbolt Hub Monitor (U2726DER)
As an affiliate, we earn on qualifying purchases.
How Cairn Panels Exposed Activity
The investigation followed a separate September report by Gambit Security, which described a campaign using three AI “harnesses”—software that enables AI models to take actions—including Cairn. Gambit reported at least 27 companies compromised and at least 600,000 payment-card records taken from two of them. That was a separate campaign; the figures should not be treated as evidence of data theft in the activity described by Reuters and NorForge.
After the Gambit report, NorForge searched for exposed dashboards associated with Cairn. Its researcher found panels that could be viewed without a password and monitored some activity through them. NorForge grouped the activity into three clusters. The largest, which it calls AIA0001, used Chinese-language task instructions and accounted for more than 1,100 attack projects across 14 panels, according to the firm.
Cairn is a general-purpose, open-source agent coordination platform rather than a tool made specifically for intrusions, according to NorForge. Its author, who uses the name le1xia0, told Reuters it “was not designed for hacking attacks” and said harmful use was not the author’s intention. The reported activity describes how operators used the software, not a claim that Cairn’s developer directed or supported the campaign.
“Most need authorization to access. Most.”
— NorForge report
Data Access and Attribution Remain Unclear
Reuters could not establish how much data, if any, was taken in the campaign. Deere told Reuters it had no indication that sensitive information was accessed. SharkNinja said a standalone product-support site was affected and that the threat had been contained. Cummins said it was investigating a potential attack on a third-party vendor’s website. These statements describe separate company responses and do not resolve whether information was accessed at other sites.
The operator’s identity has not been established publicly. The reports describe the actor as Chinese-speaking; that language finding does not, by itself, confirm the person’s nationality, location or affiliation. The source material reports no arrest or criminal charge. NorForge’s cost estimate is also uncertain, and its observations from exposed panels do not establish the full scope of activity beyond the systems it could see.
Company Reviews and Security Checks
Companies named in the reporting may continue investigating affected websites, including systems run by vendors, and determine whether any information was accessed. Cummins said it was investigating; SharkNinja said the threat had been contained; Deere said it had taken remediation actions. The reports provide no public timetable for further findings or for any law-enforcement action.
NorForge’s published findings give security teams a basis to check whether agent-platform dashboards are exposed to the internet and whether they require authentication. Further reporting may clarify the number of affected organizations, the extent of any data access and whether the operator can be identified. Until then, the documented site break-ins and company statements should be distinguished from unresolved questions about stolen data and attribution.
Key Questions
What happened in the reported campaign?
An unidentified operator used AI agents coordinated through Cairn to probe online shops. Reuters said records reviewed by its reporters indicated break-ins at websites of more than 100 companies.
Was Cairn built as a hacking tool?
No. NorForge describes Cairn as a general-purpose agent coordination platform. Its author told Reuters it was not designed for hacking attacks and said harmful use was not the author’s intention.
Was customer or company data stolen?
That has not been established for this campaign. Reuters could not determine how much data, if any, was taken. Deere said it had no indication sensitive information was accessed; other companies gave separate updates about their investigations or response.
Who was behind the activity?
The operator has not been publicly identified. NorForge described the largest activity cluster as using Chinese-language instructions, but the reporting does not establish the operator’s identity, nationality or affiliation.
How much did the campaign cost?
NorForge researcher Snorre Fagerland estimated roughly $5,000 in AI costs, or about $50 per successful intrusion. NorForge said the estimate carried considerable uncertainty, so it should not be treated as a verified accounting.
Source: ThorstenMeyerAI.com
