Signs of 'Prtseek' and 'ARTEX' use in attack materials from around 20 universities
Developer linked to Chinese hacking community identified; actual perpetrators remain unknown
Signs of login credentials and bank account information stolen by exploiting vulnerabilities in research administration and accounting systems

A line of Chinese commentary remained at the end of a file used to deface a domestic university's webpage. It said, “Prtseek is for use only in learning and authorized environments.” A warning attached to a tool designed for legitimate security checks had been found in a file linked to a university hack. This is a clue suggesting that an AI-based penetration testing tool may have been used in a series of recent cyberattacks targeting around 20 universities in South Korea.

On the 5th, a screen at the Korea Internet & Security Agency's Seoul office in Songpa-gu, Seoul, displayed a "Caution" alert for a cyber crisis in the private sector.

On the 5th, a screen at the Korea Internet & Security Agency's Seoul office in Songpa-gu, Seoul, displayed a "Caution" alert for a cyber crisis in the private sector.

View original image

On November 22, S2W, a South Korean cyber threat analysis company, said it had analyzed files released by an attacker known as “KimJEN,” who claimed responsibility for the university hacks. The company found traces of a Chinese-language AI penetration testing tool called ARTEX and Prtseek, which was developed based on ARTEX.


Prtseek is a penetration testing tool originally developed to check systems for security vulnerabilities. ARTEX, the underlying technology, is an AI agent that uses large language models (LLMs) to analyze systems and carry out the penetration process. It was also suspected of being used in a hacking incident targeting South Korean financial institutions last month, in October.


The clues found by S2W were not limited to a simple sentence in Chinese. The leaked materials contained ARTEX task paths and task identification numbers (IDs), as well as a penetration test results report believed to have been written by an AI agent. Some task logs also showed that the AI had been instructed to carry out the work under the pretense that the real system intrusion was part of an authorized penetration testing competition.


S2W also traced the development of the tool believed to have been used in the attacks. It confirmed that a user known as “Hwangdu,” who is active in Chinese hacking communities, had participated in developing Prtseek.


The Chinese-language comments and the developer's activity history shed light on the technical background of the tools used. They do not, however, identify the attacker. S2W said it had not confirmed whether Hwangdu was actually involved in the university hacks, and had not identified those behind the attacks. Some of the Chinese-language traces may also have been generated automatically by an AI tool.


KimJEN, who claimed responsibility for the hacks, said that since November 12, they had infiltrated South Korean universities and public institutions. They posted related materials on Telegram and also publicized their activities on dark web forums.


KimJEN claimed to be a European national who speaks Korean. However, some of the Korean sentences in the published conversations had unnatural structures and expressions. Cybersecurity experts have raised the possibility that a translation tool was used and are analyzing the attacker's active hours and language use, among other clues.


The actual method of intrusion appears to have focused on exploiting vulnerabilities in university administrative and accounting systems. According to S2W, the attacker exploited security vulnerabilities in rERP, an integrated research administration system used by around 20 universities, to access files inside their servers.


During this process, there were signs that authentication keys used for remote deployment and web server configuration information had been leaked. The attacker is believed to have used the access information they obtained to install a means of remotely controlling the servers and gain the privileges needed to execute commands within the systems.


Separately, signs of intrusion were found in the UNIVais accounting information systems of four universities. The attacker is believed to have accessed internal data without proper authentication and viewed or extracted login credentials, user information, and account information. Systems handling universities' research administration and financial operations were targeted.


This case is drawing attention because AI can automate tasks needed for hacking. Previously, attackers had to find security vulnerabilities, analyze intrusion paths, and collect data themselves. Misusing AI-based penetration testing tools could automate a substantial portion of this work.


This has raised concerns that even hackers without advanced technical skills could use publicly available tools to attack multiple institutions at once. Universities using the same administrative and accounting systems are particularly at risk, as common security vulnerabilities could lead to large-scale breaches.



S2W recommended that “educational institutions, including universities, review externally accessible administrator pages and deployment and testing systems, and address features that can be used without authentication.” It also stressed that “credentials and database access information that may have been leaked should be replaced.”


This content was produced with the assistance of AI translation services.

© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.

Today’s Briefing