Unauthorized access targeting dozens of companies in Japan and South Korean financial institutions was confirmed, apparently carried out by ARTEX, a China-linked AI agent that works autonomously once given instructions. A separate case was also reported in the United States.
The possible misuse of China-linked AI software in unauthorized access incidents targeting companies has emerged. The software was reportedly an AI agent, and dozens of companies, mainly in domestic industries such as retail, finance and logistics, were targeted and disclosed information leaks.
ARTEX, used in the attacks, is said to perform tasks autonomously simply by being given instructions, and traces of unauthorized access to South Korean financial institutions were also found. An attack in the United States believed to involve a different China-linked AI was reported, raising concerns that the issue could become international.
A security company under GMO Internet Group is leading the domestic investigation and, in cooperation with external organizations, identified 6 devices targeting only Japanese companies. The devices were also confirmed in connection with cases involving the use of ARTEX.
In addition, 8 companies disclosed information leaks in September and October, indicating that the impact is spreading both domestically and internationally.