Multiple reports reveal that Hong Kong has become a hub for AI cyberattacks, with over 60% of Android remote Trojan malware involving local networks. (Hao Benni)

CN
24 minutes ago

179488380187167.jpg

In recent months, several cybersecurity research reports have described Hong Kong as an important base for artificial intelligence (AI) cyber attack infrastructure. The tools used, the victims, and the operators involved in related cases vary, yet there is a repeated choice to deploy within Hong Kong networks. As AI lowers the thresholds for automated attacks, Hong Kong, as an international data and network hub, is also becoming a focus of overseas threat intelligence reports that are under continuous surveillance.

AI Penetration Testing in Hong Kong Accounts for 18%, Just Below the United States

Mobile security company Zimperium recently released a report on the Android remote control Trojan ArcticRAT, which revealed that two-thirds of the 17 command and control server IP addresses used by the malware pointed to Hong Kong.

179488380528922.jpg

The feature of ArcticRAT is its ability to quickly filter high-value content using AI, integrating xAI's Grok into the processing flow of the stolen data to identify banks, cryptocurrency, and trading applications.

The research team under software company Elastic has also recently confirmed an IP address hosted on a local network in Hong Kong running the open-source AI offensive and defensive platform CyberStrikeAI. Between September 30 and October 16, it accessed four accounts using credentials of eight Amazon AWS identities, generating 78 security alerts, and had once issued 56 requests in 141 milliseconds, aligning with the speed of automated attack tools.

Cybersecurity giant Check Point also released a report on the open-source AI penetration testing framework ARTEX earlier this month. Researchers found 1,608 publicly accessible instances, with more than half located in the United States (54%), followed by Hong Kong (18%).

The tool was previously used by attackers to breach several financial institutions in South Korea, with a CrowdStrike report designating a Hong Kong IP address as “the primary infrastructure controlled by the attackers.”

The Cases Are Increasing, Not Isolated Incidents; Authorities Should Take Them Seriously

Even as early as July this year, the Thai Ministry of Finance was breached, and the cyber threat intelligence platform Hunt.io promptly identified a group of related Hong Kong infrastructure, utilizing the Hermes AI agent to automatically enumerate hosts, search files, and analyze privilege escalation paths. Last month, incident response company Profero also disclosed a set of AI agent attack environments concentrated on networks hosted in Hong Kong, with three hosts responsible for directly launching attacks, impersonating relays, and aiming to conduct reverse engineering through AI.

The aforementioned incidents reflect that Hong Kong is becoming more than just a network transit point; it is now an operational node within the AI attack chain. As the number of cases where attack frameworks and hosts end up in locally hosted networks gradually increases, the issues may no longer be seen as isolated abuse incidents.

Relevant vendors must enhance their monitoring and response mechanisms, and authorities should not only focus on protecting local institutions but also further consider how to reduce the risk of local networks being used as platforms for cross-border attacks.

This article is reproduced from the Sing Tao Daily, with copyright owned by the original author and original media.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到[email protected],本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink