From August–September 2026, Chinese-speaking attackers used an open-source AI penetration testing framework called 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.
The attackers supplied a target and a goal, then Cairn planned, executed and recorded results on its own. These attacks achieved unauthorized access or data retrieval at 10 of the targeted organizations, including:
- Accessing dozens of valid credentials
- Mass exposing personal data (exceeding one million records at one organization)
- Retrieving military and government identities
- Creating persistent administrator accounts in a production system
- Modifying a live industry standards-ballot document
Read more…
Source: Palo Alto Unit 42
Sign up for the Cyber Security Review Newsletter
The latest cyber security news and insights delivered right to your inbox
Related:
- Ryuk Ransomware Partners with TrickBot to Gain Access to Infected Networks
January 12, 2019
Historically, Ryuk has been considered a targeted ransomware that scopes out a target, gained access via Remote Desktop Services or other direct methods, stole credentials, and then targeted high profile data and servers to extort the highest ransom amount possible. Ryuk has been a high profile ransomware due to its wide impact on the networks it infects, high ransom ...
- A Zebrocy Go Downloader
January 11, 2019
Last year at SAS2018 in Cancun, Mexico, “Masha and these Bears” included discussion of a subset of Sofacy activity and malware that we call “Zebrocy”, and predictions for the decline of SPLM/XAgent Sofacy activity coinciding with the acceleration of Zebrocy activity and innovation. Zebrocy was initially introduced as a Sofacy backdoor package in 2015, but the Zebrocy ...
- TA505 Crime Gang Debuts Brand-New ServHelper Backdoor
January 11, 2019
The latest malware from TA505 has been seen targeting banks, retailers and restaurants with two different versions. A new backdoor named ServHelper has been spotted in the wild, acting as both a remote desktop agent as well as a downloader for a RAT called FlawedGrace. According to Proofpoint, the prolific cybercriminal gang known as TA505 developed ServHelper, which has ...
- ‘Unprecedented’ DNS Hijacking Attacks Linked to Iran
January 10, 2019
The attacks, targeting several countries to redirect traffic and harvest credentials, have been linked to Iran. A wave of DNS hijacking attacks targeting victims in North America, Europe, Middle East and North Africa have been linked to Iran. The attacks, which have been ongoing over the past two years, have had “a high degree of success” ...
- New tool automates phishing attacks that bypass 2FA
January 9, 2019
A new penetration testing tool published at the start of the year by a security researcher can automate phishing attacks with an ease never seen before and can even blow through login operations for accounts protected by two-factor authentication (2FA). Named Modlishka –the English pronunciation of the Polish word for mantis– this new tool was created ...
- Ransomware MongoLock Immediately Deletes Files, Formats Backup Drives
January 8, 2019
We have been following a new wave of MongoLock ransomware attacks that immediately deletes files upon infection instead of encrypting it, and further scans for other available folders and drives for file deletion. In the wild since December 2018, the ransomware demands a payment of 0.1 bitcoin from victims within 24 hours to retrieve the ...