🔍 Read the full analysis: How AI Agents Put 100+ Company Websites In One Hacker’s Reach 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 Reuters investigation, drawing on records left exposed by the attacker, found an AI-agent operation that probed more than 1,200 websites and broke into websites belonging to more than 100 companies. The operator’s identity, the amount of data accessed, and the full impact remain unclear; several affected companies described remediation or ongoing investigation.
A Chinese-speaking operator used software agents coordinated through the open-source platform Cairn to probe online shops across several countries, with break-ins reported at more than 100 companies, Reuters reported October 19. The account draws on activity records left on the attacker’s exposed control panels; the operator’s identity and whether sensitive data was taken remain unknown.
According to Reuters and security firm NorForge, the campaign began on September 25 and targeted e-commerce websites in the United States, Canada, Australia and Japan. By mid-October, the agents had attempted to access more than 1,200 sites. Reuters counted breaches at websites associated with over 100 companies based on the attacker’s Cairn records. NorForge’s report, published October 19, grouped activity across exposed control panels and did not name victims.
Reuters independently reviewed the attacker’s Cairn activity and identified companies whose websites had been accessed. Among those named were 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 run by a third-party vendor. Reuters said it could not establish how much data, if any, was taken.
NorForge researcher Snorre Fagerland found more than 20 exposed Cairn control panels that the firm linked to likely malicious activity. The largest cluster, labeled AIA0001 by NorForge, recorded more than 1,100 attack projects across 14 panels, mostly targeting the United States, Japan and Australia. NorForge measured a little under 23 hours from start to goal for successful projects in that cluster. Its estimate of roughly $5,000 in AI costs for the campaign, or about $50 per successful intrusion, is an estimate, and the firm described its uncertainty as considerable.
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 Website Intrusions Matter
The reported operation shows how software agents can carry out prolonged, repeated reconnaissance with limited direct attention from an operator. The key change described by NorForge is not that AI has made every intrusion effortless, but that automated probing can continue for hours or days across a large list of targets. That can increase the number of websites defenders must check and the volume of activity they need to distinguish from ordinary traffic.
The reported targets also include a standalone support site and a vendor-operated website, not only a company’s main storefront. Such sites can be less central to a business’s security planning while still being connected to customer services or internal processes. The incidents make third-party and less-visible web infrastructure relevant to the assessment of a company’s exposure. The available reporting does not establish that all affected sites exposed customer data or enabled access to wider corporate networks.
NorForge’s cost figures are not independently confirmed campaign accounts, and they should not be read as a standard price for a successful attack. Even so, if automated work lowers the effort needed to keep testing potential weaknesses, organizations may face persistent activity at a scale that is difficult to review manually. The immediate practical issue is whether exposed sites, vendor systems and agent control panels are monitored and secured.
Products Worth Considering
Apple iPad Air (4th Generation, 10.9-inch, Wi-Fi, 256GB) – Space Gray
Dell 24 Plus Monitor S2425HSM – 23.8 FHD (1920×1080) 144Hz 1ms, 2 x 3W Speakers, HDMI, Height/Tilt/Pivot/Swivel Adjustability, AMD FreeSync
FLEXISPOT 36-Inch Height-Adjustable Standing Desk Converter
As an affiliate, we earn on qualifying purchases.
How Cairn Panels Exposed the Campaign
The investigation began with a separate security report. In September, Gambit Security described another agent-driven campaign after recovering a criminal staging server. It found three “AI harnesses,” or programs that let AI models act through software tools, including Cairn. Gambit reported at least 27 companies compromised and at least 600,000 card records taken from two of them. That was a separate campaign; the records do not show that those card thefts occurred in the operation Reuters later described.
NorForge then searched for internet-accessible Cairn dashboards. The firm said most panels require authorization, but it found some with no password protection. From around October 2, researcher Snorre Fagerland identified roughly two dozen panels being used for malicious activity, according to Reuters. NorForge sorted the activity into three clusters and published its findings on October 19. The exposed dashboards gave researchers a view into tasks and results that would otherwise have been harder to verify.
Cairn is a general-purpose, open-source agent platform, not a product designed specifically for cyberattacks. It coordinates agents through a shared record of information found and planned next steps. Its creator, who uses the name le1xia0, told Reuters Cairn “was not designed for hacking attacks.” The platform’s stated purpose and the way a user applies it are separate questions: in this case, researchers say the records showed it being used to automate website intrusions.
““Necessary remediation actions” were taken, and the company had “no indication that sensitive information was accessed.””
— Deere, in a statement to Reuters
What Investigators Still Cannot Confirm
The operator has not been publicly identified. The source material describes the person behind the activity as a Chinese-speaking actor; it does not establish their name, location or affiliation. It also reports no arrest or criminal charge. The presence of exposed control panels offered researchers a view of the campaign, but does not resolve who directed it.
The amount and type of information accessed remain unclear. Reuters could not determine how much data, if any, was taken, and the company statements cited so far do not establish the full scope of each intrusion. Nor is it clear whether access was used beyond the websites described, whether it was sold or monetized, or how many of the more than 100 reported company websites contained sensitive information. NorForge characterized the apparent goal as gaining persistent access to e-commerce or payment infrastructure for later monetization; that is the firm’s interpretation of the activity, not proof that such monetization occurred.
The cost and efficiency figures also carry qualifications. NorForge estimated roughly $5,000 in AI expenses and about $50 per successful intrusion, while warning of considerable uncertainty. Those figures are not a verified accounting of the operator’s spending and may not represent the costs of other campaigns.
Company Reviews and Security Follow-Up
The next developments are likely to come from affected companies’ investigations and any further reporting based on the attacker’s records. Deere has described remediation and said it had no indication that sensitive information was accessed; SharkNinja said it contained the threat; Cummins said its investigation was continuing. The available reports do not give a common deadline for those reviews or say when their findings will be public.
Security teams will need to determine whether affected websites, vendor systems and related credentials have been secured, and whether there is evidence of data access beyond the initial intrusion. Researchers may also continue monitoring internet-exposed Cairn panels. Until companies or investigators release more detail, the confirmed scope of data exposure and downstream effects remains unresolved.
Key Questions
What happened in the reported campaign?
Reuters reported that an operator used Cairn-coordinated AI agents to probe more than 1,200 websites and gain access to websites associated with over 100 companies. The count was based on activity records left on exposed attacker control panels.
Which companies were named?
Reuters identified Deere, SharkNinja and Cummins among the companies affected. Their statements describe remediation, a contained threat at a standalone support site, and an investigation involving a vendor-operated website, respectively.
Was customer or company data stolen?
That has not been established for the campaign described by Reuters. The investigation could not determine how much data, if any, was taken. Deere said it had no indication that sensitive information was accessed; the full impact across all affected websites remains unclear.
Who operated the AI agents?
The operator has not been publicly identified. The reporting describes the actor as Chinese-speaking, but does not confirm their identity, location or affiliation.
Was Cairn built for hacking?
No. Cairn is described as an open-source, general-purpose agent platform. Its creator told Reuters it was not designed for hacking attacks, while researchers say the campaign records showed it being used to automate intrusions.
Source: ThorstenMeyerAI.com
