Hacking used to demand time, patience and a great deal of specialist knowledge. According to a new Japanese-led investigation, artificial intelligence is quickly changing that equation. The Japan Computer Emergency Response Team Coordination Center (JPCERT/CC) said on Saturday that around 1,000 servers suspected of serving as bases for AI-driven cyberattacks have been online since October.
The probe, which JPCERT/CC began in early October together with overseas cybersecurity firms, found that the machines may have used AI to hunt for weaknesses and target websites around the world. The center says it is currently tracking 600 to 700 of the servers.
Most of them hosted AI-powered platforms such as the penetration-testing tool ARTEX, a system released in Chinese in July that uses multiple AI agents to find vulnerabilities automatically. The servers appeared to be scanning for security holes and trying to steal configuration files. One machine even contained traces of an AI coding assistant from the US company Anthropic, a sign that attackers are turning legitimate developer tools into weapons.
Investigators also discovered software for dissecting the inner workings of smartphone apps, along with programs for probing system vulnerabilities. The attackers are believed to have analysed apps and then tried to break in through the interfaces those apps use to exchange data with outside systems, so-called APIs.
The findings build on earlier warnings. In October, JPCERT/CC and the security firm GMO Cyber Security by IERAE reported evidence of AI-assisted attacks on Japanese companies, describing attacker servers that worked like a well-stocked “toolbox”.
For businesses, the message is clear: patch quickly, lock down APIs and watch for automated scanning. As JPCERT/CC analyst Ken Sajo put it, AI is still in the early stages of being used for cyberattacks, and “more sophisticated attacks will increase in the future.”





