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:
- Threats and Consequences A Security Analysis of Smart Manufacturing Systems
May 11, 2020
In the era of Industry 4.0, there has been increasing adoption of smart manufacturing technologies by organizations looking to improve their manufacturing efficiency. While this has provided plenty of benefits, such as enhanced productivity at lower costs, it has also introduced new attack vectors that can be exploited by threat actors looking to gain a foothold ...
- Zeus Sphinx revamped as coronavirus relief payment attack wave continues
May 11, 2020
The Zeus Sphinx banking Trojan is now receiving frequent updates and upgrades to its malicious arsenal while being deployed in active coronavirus scams. On Monday, IBM Security researcher Nir Shwarts said the company has been tracking the evolution of the malware which is based on the leaked codebase of the well-known Zeus v.2 Trojan. Zeus Sphinx — also referred ...
- Thunderbolt flaws affect millions of computers – even locking unattended devices won’t help
May 11, 2020
A Dutch researcher has detailed nine attack scenarios that work against all computers with Thunderbolt shipped since 2011 and which allow an attacker with physical access to quickly steal data from encrypted drives and memory. Researcher Björn Ruytenberg detailed the so-called Thunderspy attacks in a report published on Sunday, warning that the attacks work even when users ...
- North Korean hackers infect real 2FA app to compromise Macs
May 9, 2020
Hackers have hidden malware in a legitimate two-factor authentication (2FA) app for macOS to distribute Dacls, a remote access trojan associated with the North Korean Lazarus group. Dacls has been used to target Windows and Linux platforms and the recently discovered RAT variant for macOS borrows from them much of the functionality and code. The threat actor planted the ...
- A hacker group is selling more than 73 million user records on the dark web
May 9, 2020
A hacker group going by the name of ShinyHunters claims to have breached ten companies and is currently selling their respective user databases on a dark web marketplace for illegal products. The hackers ardarke the same group who breached last week Tokopedia, Indonesia’s largest online store. Hackers initially leaked 15 million user records online, for free, but ...
- Threat Brief: Maze Ransomware Activities
May 8, 2020
Since the beginning of the calendar year, Palo Alto Networks has detected an uptick in Maze ransomware samples across multiple industries. As a result, we’ve created this general threat assessment post on the Maze ransomware activities and a full visualization of these techniques can be viewed in the Unit 42 Playbook Viewer. Maze ransomware, a variant of ChaCha ...