Unit 42 says attackers used Cairn AI against 19 organizations
Original title: Palo Alto tracks a massive Chinese AI attack that sadly lives up to industry fears
Access succeeded at 10 targets; one business lost more than one million records, Unit 42 reports.
Palo Alto Networks Unit 42 says it tracked a China-based campaign in August and September 2026 that used a customized fork of the open-source AI penetration-testing framework Cairn with Zhipu GLM 5.2 and DeepSeek V4 Flash/Pro. It reports 25 operations against at least 19 organizations, with unauthorized access at 10, dozens of valid credentials, persistent admin accounts and, through separate flaws, customer-level read, write and delete access. One business had more than one million records captured; elsewhere military and government identities were taken and a live standards ballot was overwritten. Unit 42 says operators ran from Hong Kong with a US egress node and believes one person or a handful conducted the activity.
Unconfirmed: Unit 42 attributes the activity to Chinese-speaking attackers and believes few people ran it; targeted organizations are not named.
- Operations lasted from 20 minutes to more than 50 hours, with three to seven stages.
- Cairn used a shared blackboard; humans could inject hints during operations.
- Each task was seeded with a root domain and an authorization.md pretext file.
- Unit 42 says stock Cairn prompts already tell the model never to refuse a task.
- A researcher said CISOs should watch unauthorized changes to critical records.
Impact: Unauthorized access, credential theft and altered records
- 1.Monitor unauthorized changes to critical records and privileged access.
- 2.Test ability to verify and restore trusted information after an incident.
- ATT&CK
- T1190 Exploit Public-Facing ApplicationT1078 Valid AccountsT1136 Create AccountT1565.001 Stored Data Manipulation
- Malware
- Cairn
- Organizations
- Palo Alto NetworksUnit 42
- Countries
- ChinaHong KongUnited States
From the original article
4 min read at csoonline.com
Palo Alto Networks Unit 42 is reporting today that it has tracked a massive China-based attack leveraging the open-source AI penetration testing framework Cairn to autonomously conduct 25 operations against at least 19 organizations. The attackers weaponized Cairn by pairing it with Chinese AI models Zhipu GLM and DeepSeek, Unit 42 said.
csoonline.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.