🔍 Read the full analysis: What The Cost Of AI-Assisted Website Break-Ins Could Mean on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Reuters, drawing on an attacker’s exposed Cairn records, reported that an AI-assisted campaign tried more than 1,200 websites and broke into sites belonging to more than 100 companies. A security researcher estimated the campaign’s AI costs at about $5,000, but called the estimate highly uncertain; the amount of data accessed or taken has not been established.
An operator used AI agents coordinated through Cairn to target online shops, with the attacker’s records indicating break-ins at websites of more than 100 companies, Reuters reported October 19. A security researcher estimated the campaign’s AI costs at about $5,000, though the estimate is uncertain; the extent of any data access or theft is not known.
According to Reuters and the security firm NorForge, the campaign began on September 25 and targeted online shops in the United States, Canada, Australia and Japan. By mid-October, the agents had tried more than 1,200 sites. Reuters identified affected company websites by reviewing records in the attacker’s Cairn activity and contacting the companies. The reported victims include farm-equipment maker Deere, appliance maker SharkNinja and engine maker Cummins.
NorForge researcher Snorre Fagerland found exposed Cairn dashboards while searching for publicly accessible agent-control panels. NorForge linked more than 20 panels to likely malicious activity and grouped the activity into three clusters. Its largest cluster, which it labels AIA0001, used Chinese-language task instructions. NorForge counted more than 1,100 attack projects across 14 panels; these are projects, not a confirmed count of successful intrusions.
NorForge measured successful AIA0001 projects as taking a little under 23 hours from start to goal, on average as described in its report. Fagerland estimated roughly $5,000 in AI costs for the campaign, or around $50 per successful hack. NorForge cautioned that the cost estimate carries considerable uncertainty. The source material does not provide an independently verified accounting of the campaign’s total expenses.
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 Low-Cost Automation Matters
The reported costs point to a potential shift in how attackers allocate effort: software agents can keep testing targets over many hours without an operator directing every step. If the estimate is broadly accurate, the AI portion of this campaign may have been inexpensive relative to the number of sites targeted. But $5,000 is a researcher’s estimate, not a verified invoice or a measure of all costs, and it does not establish the overall financial impact on victims.
The report also focuses attention on websites that may receive less security scrutiny than a company’s main online services. NorForge’s findings describe attacks on e-commerce infrastructure, while affected companies’ comments included references to a standalone support site and a site operated by a third-party vendor. That does not prove that every small or vendor-operated site is at heightened risk, but it illustrates why organizations may need to account for less prominent services and outside providers when reviewing exposure.
Automation can also increase the amount of activity defenders must assess. A high volume of attempts is not the same as a high number of successful intrusions, and the reporting does not quantify how much extra work the campaign created for security teams. Still, the reported scale makes monitoring and prioritizing alerts relevant concerns, especially where multiple agents can pursue tasks in parallel.
Products Worth Considering
Samsung 27-Inch Odyssey QD-OLED G8 4K 240Hz Gaming Monitor
Samsung 27-inch Odyssey OLED G6 QD-OLED QHD Gaming Monitor
Samsung 27 Odyssey OLED G5 (G50SF) QHD QD-OLED Gaming Monitor
As an affiliate, we earn on qualifying purchases.
How Researchers Found the Campaign
The reporting builds on a separate investigation by Gambit Security in September. Gambit said it recovered a criminal staging server containing three AI harnesses—software that allows AI models to take steps toward a goal—including Cairn. Its report described a distinct agent-driven campaign that compromised at least 27 companies and took at least 600,000 card records from two. Those figures relate to Gambit’s separate investigation, not the campaign described by Reuters and NorForge.
After Gambit’s report, NorForge searched for exposed dashboards belonging to Cairn, an open-source agent platform. Fagerland found control panels that were accessible without a password and could observe activity on some of them. NorForge said it identified roughly two dozen panels associated with malicious use from around October 2, with additional panels appearing later. The firm published its report on October 19; Reuters published its investigation the same day.
Cairn is not described by NorForge as a hacking-only product. The platform’s author, who uses the name le1xia0, said it was not designed for hacking attacks and that harmful use was not their intention. The investigation therefore concerns how an open-source agent system was used by an unidentified operator, not evidence that the platform’s author directed the activity.
“a little under 23 hours”
— NorForge report
What the Records Cannot Establish
The identity of the operator has not been established. NorForge described the actor as Chinese-speaking, but that does not confirm the person’s identity, location or nationality. The reporting also does not say that anyone was arrested or charged.
It remains unclear how much data, if any, was accessed or taken in the campaign Reuters investigated. 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 was contained. Cummins said it was investigating a potential cyberattack on a website operated by a third-party vendor. These company statements describe their own responses; they do not settle what happened across all reported sites.
Reuters could not establish how much data, if any, was taken. The count of more than 100 affected company websites comes from Reuters’ review of the attacker’s records and company outreach; NorForge’s report itself did not name victims. The cost estimate is likewise not independently confirmed, and the reported number of targeted sites should not be confused with the number of successful break-ins.
Company Reviews and Further Findings
The affected companies’ investigations and disclosures may clarify whether information was accessed, what systems were involved and whether third-party services played a role. Cummins said it was investigating a potential attack on a vendor-operated website. Deere and SharkNinja gave Reuters statements about their response, but the source material does not report a complete account of the incident from each company.
Further findings could also test the limits of the available evidence: the exposed Cairn panels gave researchers an unusually direct view of attacker activity, but they do not by themselves provide a complete, independently verified record of every target, outcome or cost. At present, the scale of targeting is clearer than the campaign’s full impact. No additional investigation deadline or law-enforcement milestone is specified in the reports.
Key Questions
How many websites did the AI agents target?
Reuters reported that the agents tried more than 1,200 sites by mid-October, based on the attacker’s records. That is a count of attempted targets, not successful intrusions.
How much did the campaign cost?
NorForge researcher Snorre Fagerland estimated about $5,000 in AI costs, or roughly $50 per successful hack. NorForge described the estimate as carrying considerable uncertainty, and the figure is not a verified total campaign cost.
Was sensitive information stolen?
The available reporting does not establish how much data, if any, was taken. Deere said it had no indication that sensitive information was accessed; Reuters could not determine the amount of data accessed or taken across the reported incidents.
Who operated the agents?
The operator’s identity is unknown. NorForge described the actor as Chinese-speaking, but that does not establish who they are or where they were located. The source material reports no arrest or charge.
Was Cairn built for cyberattacks?
No. NorForge described Cairn as a general-purpose, open-source agent platform. Its author said it was not designed for hacking attacks and that harmful use was not their intention; the reporting concerns the platform’s use by an unidentified operator.
Source: ThorstenMeyerAI.com
