Dark ReadingElizabeth Montalbano
China linked actors use AI framework to breach 10 organisations
Part of the CL-STA-1238 story · since 16 Nov
- Mon 16 Nov, 18:00 UTC. Dark Reading, first report. China linked actors use AI framework to breach 10 organisations. This article, 1st of 1.
China-linked Nexus actors used the Cairn AI penetration-testing framework to run autonomous, AI-driven attacks against at least 19 organisations from August to September 2026, with unauthorized access or data retrieval achieved at 10 of them. Researchers at Palo Alto Networks’ Unit 42 track the operation as CL-STA-1238, and describe Cairn as planning, executing and recording results without constant human direction.
The attacks targeted advanced manufacturing and defence sectors, as well as civil society groups and human rights organisations. Among the confirmed outcomes were dozens of valid credentials obtained, exposure of personal data exceeding 1 million records at a single victim, retrieval of military and government identities, creation of persistent administrator accounts in production systems, and modification of a live industry standards-ballot document. Seven of the 10 affected organisations were in the semiconductor and advanced-manufacturing supply chain, with three in US defence or defence policy.
Cairn was deployed using Chinese large language models Zhipu GLM 5.2 and DeepSeek V4 Flash and Pro, illustrating a shift in how AI is weaponised for targeted operations. Initial access came through familiar security gaps across industries, according to Unit 42, including authentication bypasses; insecure direct object references; mass-assignment privilege escalation; weak and default credentials; exposed cloud APIs; and dated public CVEs.
The campaign demonstrates that attackers can operate multiple Cairn “agents” in parallel, each conducting reconnaissance, vulnerability identification and exploitation, and reporting back via a shared blackboard, while human operators can supply hints when needed.
The report notes that traditional security practices—identity hygiene, attack-surface reduction and software patching—remain relevant, alongside enhanced continuous attack-surface management and cyber deception to detect reconnaissance and lateral movement before objectives are achieved.
How the story developed
1 article from 1 outlet since 16 Nov. Still developing; the latest report was 6 hours ago. This was the first report.
- Mon 16 Nov, 18:00 UTC. Dark Reading, first report. China linked actors use AI framework to breach 10 organisations. This article, 1st of 1.
- No CVE
- Larger: an outlet’s first report
- Coverage over time
Show all 1 articles as a listHide the list
- Mon 16 NovDay 01
- China linked actors use AI framework to breach 10 organisations18:00 · Dark Reading · First report · This article, 1st of 1