ARTEX Closure Turns South Korea Bank Breaches into an AI Governance Test

Seoul, 9 October 2026 – The developer of ARTEX has stopped public updates and moved the artificial-intelligence penetration-testing agent to a closed-source model after security researchers linked the tool to attacks on South Korean financial institutions. The decision turns a bank-cybersecurity incident into a broader governance question for investors: how should powerful dual-use AI systems be distributed when legitimate testing functions can also accelerate criminal intrusion?

The developer, who uses the online handle Autumn-27, said ARTEX was created to help enterprises identify security weaknesses and opposed illegal use. The project will no longer release new public versions or provide maintenance support. That response may limit casual access to future capabilities, but it cannot recall existing code or models already downloaded, leaving financial institutions to manage the continuing risk from tools in circulation.