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:
- Newly-Discovered Malware Targets Unpatched MacOS Flaw
June 25, 2019
Researchers have discovered never-before-seen Mac malware samples, which they believe are being developed to target a recently-disclosed vulnerability in the MacOS operating system. The vulnerability, a bypass that was disclosed in May and has yet to be patched by Apple, exists in the MacOS Gatekeeper security feature, which verifies downloaded applications before allowing them to run on Macs. ...
- Riltok mobile Trojan: A banker with global reach
June 25, 2019
Riltok is one of numerous families of mobile banking Trojans with standard (for such malware) functions and distribution methods. Originally intended to target the Russian audience, the banker was later adapted, with minimal modifications, for the European “market.” The bulk of its victims (more than 90%) reside in Russia, with France in second place (4%). ...
- Anonymous hacker exposed after dropping USB drive while throwing Molotov cocktail
June 24, 2019
In a bizarre investigation, Belgium police have identified a member of the Anonymous Belgium hacker collective while investigating an arson case at a local bank. The perpetrator, a 35-year-old man from the Belgian city of Roeselare, was initially arrested after throwing a Molotov cocktail at the Crelan Bank office in Rumbeke, a suburb of Roeselare, back ...
- Consumers Urged to Junk Insecure IoT Devices
June 18, 2019
More than 2 million connected security cameras, baby monitors and other IoT devices have serious vulnerabilities that have been publicly disclosed for more than two months – yet they are still without a patch or even any vendor response. Security researcher Paul Marrapese, who disclosed the flaws in April and has yet to hear back from any impacted ...
- Mobile Cyberespionage Campaign ‘Bouncing Golf’ Affects Middle East
June 18, 2019
We uncovered a cyberespionage campaign targeting Middle Eastern countries. We named this campaign “Bouncing Golf” based on the malware’s code in the package named “golf.” The malware involved, which Trend Micro detects as AndroidOS_GolfSpy.HRX, is notable for its wide range of cyberespionage capabilities. Malicious codes are embedded in apps that the operators repackaged from legitimate ...
- Plurox: Modular backdoor
June 18, 2019
In February this year, a curious backdoor passed across our virtual desk. The analysis showed the malware to have a few quite unpleasant features. It can spread itself over a local network via an exploit, provide access to the attacked network, and install miners and other malicious software on victim computers. What’s more, the backdoor ...