Recent hacking attacks targeting domestic financial institutions have led to the reintroduction of the artificial intelligence (AI) penetration testing tool 'ARTEX' as a commercial service through a Chinese security company.
On October 16, the security industry reported that Xu Feng, co-founder and CEO of the Chinese security firm Threatbook, announced via social media the upcoming launch of a cloud platform called 'RedAgents.ai,' which will integrate multiple AI penetration testing agents, including a commercial version of ARTEX.
ARTEX is designed as a Chinese-language AI penetration testing program based on a large language model (LLM), enabling AI agents to autonomously analyze targets and execute security tools from scenario development to execution.
Between late September and early October, approximately 100,000 records were leaked from 16 domestic financial institutions, including Shinhan, KB Kookmin, Hana, and Busan Bank. Evidence indicated that ARTEX was exploited during these attacks, which occurred on October 10, with the report to authorities filed on the same day. The Cyber Investigation Bureau of the National Police Agency is currently investigating the incident.
Initially, ARTEX was publicly available as open-source on the developer platform GitHub. However, due to rising concerns over its potential misuse in cybercrime, the developer switched the repository to private and announced the project's suspension.
Investigations revealed that the developer of the program was a threat detection engineer at Threatbook, who had originally developed ARTEX as a personal project. Threatbook has since tasked this engineer with developing a commercial version called 'ARTEX PRO,' which will be a core service of the new cloud platform.
According to CEO Xu Feng, 'RedAgents.ai' will operate in a cloud environment utilizing 'Zero Data Retention' (ZDR) technology, ensuring that user data is not stored.
In addition to ARTEX PRO, the platform will feature various AI security tools, including 'YanXun,' developed by a company affiliated with Peking University, 'YanJi,' which won first place in the Tencent AI penetration testing competition, and the commercial version of 'CyberStrikeAI,' recently acquired by Threatbook.
Threatbook plans to continuously add agents that integrate with the platform in the future.
CEO Xu Feng stated, 'A single agent cannot find all vulnerabilities. Just like human-led penetration testing, AI agents have different strengths and approaches, so using multiple agents together allows for a more comprehensive security assessment.'
Threatbook is expected to release a preview version of 'RedAgents.ai' soon and is currently accepting sign-ups for early users.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.