AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: One Hacker, AI Agents, And The Changing Cost Of A Website Break-In on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on monitors, keyboards and dev gear

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

TL;DR

A Chinese-speaking operator used Cairn, a free, open-source AI-agent platform, in a campaign that probed more than 1,200 websites and was linked by Reuters to break-ins at sites associated with more than 100 companies. NorForge estimated the campaign’s AI costs at about $5,000, but that figure is uncertain, and the amount of data accessed or taken has not been established.

A Chinese-speaking operator used AI agents running on the free, open-source platform Cairn to probe more than 1,200 online shops from September 25 into mid-October, and Reuters identified website break-ins involving more than 100 companies, including Deere, SharkNinja and Cummins. The reporting draws on exposed attacker control panels and company responses; it remains unclear how much data, if any, was accessed or taken.

Security researcher Snorre Fagerland of NorForge found Cairn control panels accessible on the internet, including panels with no password protection. NorForge says it identified more than 20 panels it linked to likely malicious use. Fagerland was able to observe activity through the exposed interfaces, and NorForge grouped the activity into three clusters. Its report, published October 19, describes the largest cluster as AIA0001.

NorForge counted more than 1,100 attack projects across 14 panels in that cluster, aimed mainly at the United States, Japan and Australia. Reuters separately reported that the operator’s records showed attempts against more than 1,200 sites and break-ins at websites of more than 100 companies. Reuters said it reviewed the attacker’s Cairn activity and contacted companies; two smaller businesses acknowledged being compromised, while other contacted victims did not respond.

NorForge measured successful AIA0001 projects as taking a little under 23 hours from start to goal. It estimated about $5,000 in AI costs for the campaign, or roughly $50 per successful intrusion, while warning that the estimate carries considerable uncertainty. Those figures are estimates from the research, not independently confirmed costs. 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 contained. Cummins said it was investigating a “potential cyber-attack” on a site run by a third-party vendor.

At a glance
reportWhen: Campaign began September 25, 2026; repo…
The developmentReuters and security firm NorForge report that an operator used AI agents to automate a campaign against online shops, with Reuters identifying more than 100 affected companies.
Thorsten Meyer AI
Cluster AIA0001 · Sep–Oct 2026

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.

<100Chinese characters of instructions from the operator (Reuters)
1,200+sites the agents had tried by mid-October (Reuters)
100+companies whose websites were broken into, per Reuters' count from the attacker's records
~23haverage from start to goal for successful projects (NorForge)

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

  1. Shop websites are ranked by traffic and checked automatically in a browser
  2. Each target becomes a Cairn project: a starting point and a goal
  3. Agents post facts and next steps on a shared board and build on each other
  4. 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

Lower Costs, More Automated Probing

The reporting points to a practical change in the economics of website attacks: one operator can coordinate multiple agents to keep testing targets over extended periods, rather than manually directing each step. NorForge’s estimated $5,000 campaign cost suggests automation may make prolonged probing less expensive, although the firm itself stresses that the estimate is uncertain and should not be treated as a fixed cost for future attacks.

The affected sites also show why organizations may need to look beyond their main customer-facing platforms. The companies’ statements refer to a product-support site and a website operated by a third-party vendor. A less prominent site can still provide access or become a route into systems that matter. The reports do not establish that any such route led to sensitive corporate data, but they underline the need to account for external sites and vendor-managed services in security monitoring.

Automation also raises a workload issue for defenders. A swarm of agents can generate many actions and findings for security teams to assess. That does not mean every automated probe succeeds, or that every alert signals a breach. It does mean organizations may face more persistent activity and need reliable ways to distinguish harmless traffic from attempts to gain or retain access.

Products Worth Considering

LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1

LG UltraGear 27-Inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDMI 2.1

As an affiliate, we earn on qualifying purchases.

How Researchers Found Cairn Panels

The campaign came to wider attention after a separate investigation by Gambit Security in September. Gambit said it recovered a criminal staging server containing three AI-agent “harnesses”: Hermes, Strix and Cairn. A harness provides a way for AI models to carry out tasks through software. NorForge then searched for exposed dashboards used to manage Cairn and found panels that could be accessed without a password.

NorForge describes Cairn as a general-purpose, state-based problem-solving engine rather than a tool created specifically for hacking. In the observed campaign, agents shared information about what they had found and what they planned to do next, while the operator supplied goals and occasional guidance. The source material says the agents were directed toward online shops; this account avoids publishing operational steps that could help reproduce intrusions.

The September Gambit report described a separate agent-driven campaign, not the same set of attacks. Gambit reported at least 27 companies compromised and at least 600,000 payment-card records taken from two of them. That case provides relevant background on criminal use of agents, but its reported victims and stolen records should not be attributed to the Cairn campaign described by Reuters and NorForge.

““a little under 23 hours””

— NorForge report

Data Access and Costs Still Unknown

Reuters could not establish how much data, if any, was taken in the reported break-ins. Deere said it had no indication that sensitive information was accessed, while SharkNinja said it contained the threat on a standalone support site. Cummins described its case as a potential attack on a third-party vendor’s website and said it was investigating. These statements do not settle whether data was accessed across the wider group of reported victims.

The operator’s identity has not been established. The source reporting describes the actor as Chinese-speaking; it does not identify a person, confirm their location or report an arrest or criminal charge. NorForge’s activity counts, Reuters’ company tally and the estimated AI expenditure also have different bases. The cost figure is explicitly an estimate, and the reporting does not establish that every attempted site was compromised.

Company Investigations Continue

The companies that acknowledged incidents will need to complete or continue their investigations, including determining whether any information was accessed and whether affected sites have been secured. Cummins said it was investigating; the reporting provided no final finding from that inquiry. Reuters also may be able to clarify the scope of the incidents as more companies respond or release information.

For security researchers and organizations, the exposed Cairn panels and the reported use of agents are likely to remain areas of attention. What is not yet known is whether the same operator will continue the campaign, whether other organizations have been affected without detection, or whether the observed cost and pace will recur in other attacks. Further reporting should distinguish confirmed incidents from attempted probes and keep estimates separate from verified losses.

Key Questions

What happened in the reported campaign?

An operator used Cairn AI agents to probe online shops. Reuters reported attempts against more than 1,200 sites and break-ins involving websites of more than 100 companies, based on its review of the attacker’s records.

Who was behind the activity?

The operator has not been identified. The reporting describes the actor as Chinese-speaking, but does not establish a person’s identity or location and reports no arrest or charge.

Were customer or company records 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; the other cited company statements describe containment or an ongoing investigation.

How much did the campaign cost to run?

NorForge estimated about $5,000 in AI costs, or roughly $50 per successful intrusion. The firm characterized the uncertainty as considerable, so the numbers are estimates rather than verified expenses or a general cost benchmark.

Was Cairn created as a hacking tool?

NorForge describes Cairn as a general-purpose problem-solving platform. Its author, using the name le1xia0, told Reuters it was not designed for hacking attacks and that harmful use was not their intention.

Source: ThorstenMeyerAI.com

You May Also Like

ACE Robotics Releases A Robot-Focused Model For Spatial Awareness

ACE Robotics has released a model for robot spatial perception, but its name, capabilities, performance and availability have not been disclosed in the available report.

How Is Ropes & Gray Bringing AI Into Transactional Diligence?

An OpenAI page headline identifies Ropes & Gray’s AI work on transactional diligence, but the available extract gives no tool, timeline or results.

The Revised Rules Of Engineering Management With Will Larson

Will Larson discusses shifting expectations for engineering managers, the changing path to CTO, and how AI may bring technical decisions closer to leadership.

Software Testing Tools: A Prime Big Deal Days Guide

Compare software testing tools by purpose, cost, and upkeep, then build a dependable test setup that fits your team.