🔍 Read the full analysis: How Much Changes When A Website Break-In Costs About $50? on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
Reuters reported that an operator used Cairn, an open-source AI-agent platform, in a campaign that targeted more than 1,200 sites and broke into websites associated with more than 100 companies. NorForge estimated the campaign cost about $5,000 in AI use, or roughly $50 per successful intrusion, but said that estimate carries considerable uncertainty; the amount of data accessed or taken is not known.
AI agents were used in a campaign that targeted more than 1,200 websites and broke into sites associated with more than 100 companies, Reuters reported on October 19, citing the attacker’s own Cairn activity records. Security firm NorForge estimated the campaign’s AI costs at about $5,000—roughly $50 per successful intrusion—while warning that the estimate is uncertain. The operator’s identity and whether sensitive data was taken remain unclear.
The activity began on September 25 and targeted online shops in the United States, Canada, Australia and Japan, according to Reuters and NorForge. Reuters said it identified the affected companies by independently reviewing records in the operator’s Cairn system and contacting businesses. The named companies included Deere, SharkNinja and Cummins; the reports do not establish that every targeted site was breached.
NorForge researcher Snorre Fagerland found exposed Cairn control panels and monitored activity through them. NorForge linked more than 20 panels to likely malicious use and grouped activity into three clusters. Its largest cluster, which it called AIA0001, recorded more than 1,100 attack projects across 14 panels. NorForge said successful projects took a little under 23 hours on average and involved about 140 recorded “facts” each. These are measurements from the observed activity, not a general benchmark for AI-led cyberattacks.
The estimated $50 per intrusion is a calculation based on NorForge’s estimate of total AI costs and the number of successful attacks; the firm described the uncertainty as “considerable.” It is not a measure of the full cost of the operation, which may include infrastructure and other expenses. Reuters could not determine how much data, if any, was taken from the affected sites.
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
Low-Cost Automation, Wider Exposure
The reported cost estimate highlights how automation may reduce the expense of repeatedly probing websites. An operator can set a goal and let agents work through a sequence of tasks, reducing the need to direct every step manually. That can make it easier to sustain activity across many targets, although this case does not establish that AI agents independently chose targets or operated without human oversight.
For companies, the reported campaign also points to the risk of less prominent online properties, including product-support sites and services run by third-party vendors. A company’s main website is not the only possible entry point to investigate. At the same time, the reports do not establish that every compromised site exposed customer or payment data, or that the campaign caused wider network breaches.
Automated activity can also increase the volume of alerts and investigation work for security teams. The practical concern is not that a single cost estimate predicts future attacks, but that inexpensive, persistent probing may make it harder to separate serious incidents from routine background activity. Defenders will need to verify findings and prioritize risks without treating every automated alert as proof of a successful breach.
Popular right now
Roborock F25 BX Cordless Wet Dry Vacuum Mop
Dyson Car+Boat Cordless Handheld Vacuum
Roborock F25 GT Set Cordless Wet Dry Vacuum Cleaner and Mop with 2 Brushes
As an affiliate, we earn on qualifying purchases.
How Researchers Traced Cairn Activity
The investigation followed an earlier report by Gambit Security in September. Gambit said it recovered a criminal staging server containing three tools that allow AI models to act through software harnesses, including Cairn. It documented a separate agent-driven campaign involving at least 27 compromised companies and said at least 600,000 card records were taken from two of them. That was a different campaign; its reported thefts should not be attributed to the activity Reuters and NorForge examined.
NorForge searched for exposed Cairn dashboards, which can show tasks, findings and activity in progress. Fagerland found panels that were accessible without a password, then tracked clusters of activity and reported the findings on October 19. Reuters conducted a separate review of Cairn records and contacted companies to identify possible victims. NorForge’s report did not name victims; the company names in this account come from Reuters’ reporting.
Cairn is an open-source, general-purpose agent platform, not a hacking product by design. Its creator, who uses the name le1xia0, said it was not designed for hacking attacks and that harmful use was not the creator’s intention, according to Reuters. The platform’s involvement does not by itself establish that its developer knew about or supported the campaign.
““a little under 23 hours””
— NorForge report
Data Access and Operator Identity
The operator has not been publicly identified. NorForge described the person behind the largest activity cluster as a Chinese-speaking actor; that description does not establish the operator’s nationality or location. The source material reports no arrest or criminal charge.
The amount and type of information accessed or taken from the affected websites are also unknown. Deere told Reuters it 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 cyberattack on a website operated by a third-party vendor. These company statements describe their own responses; they do not resolve what happened across all reported sites.
NorForge’s cost estimate is not an independently audited accounting of the operation. Its calculation concerns estimated AI use and is not a confirmed total for all operational expenses. The reports also do not show whether the access was used for fraud, data theft or resale, or whether the operator maintained access after discovery.
Company Reviews and Follow-Up
The immediate next step for affected organizations is to complete their investigations, remove unauthorized access and determine whether any data was exposed. Deere, SharkNinja and Cummins have described remediation or ongoing review to Reuters, but the reports provide no common timeline for when those inquiries will conclude.
Security researchers and companies may also continue reviewing exposed Cairn panels and related records to establish the campaign’s scope. Further public findings could clarify how many of the more than 1,200 targeted sites were successfully accessed, what information was reachable and whether the operator attempted to monetize that access. No further milestone or law-enforcement action is confirmed in the source material.
Key Questions
What does the $50 figure represent?
It is NorForge’s estimate of AI costs per successful intrusion in the reported campaign, based on an estimated total of about $5,000. NorForge said the estimate carries considerable uncertainty and it is not a full accounting of operational costs.
Were more than 1,200 companies breached?
No. The reports say the agents targeted more than 1,200 sites. Reuters reported break-ins at websites associated with more than 100 companies. The number targeted is not the number confirmed as successfully compromised.
Was Cairn built as a hacking tool?
Not according to the platform’s creator, who told Reuters Cairn was not designed for hacking attacks. The reporting describes it as an open-source, general-purpose agent platform that was used in the campaign.
Was customer or payment data stolen?
That remains unclear for the campaign described by Reuters and NorForge. Reuters could not establish how much data, if any, was taken. A separate September campaign reported by Gambit Security involved card records, but it was not the same activity.
Who was behind the campaign?
The operator has not been publicly identified. NorForge described the largest cluster’s activity as coming from a Chinese-speaking actor, which does not establish the person’s identity, nationality or location.
Source: ThorstenMeyerAI.com
