South Korea's government says it is highly likely that an AI security-testing tool called Artex was used in data breaches at more than seven financial institutions, per AFP via Tech Xplore. The Financial Services Commission puts the number of people affected above 68,000. Shinhan Bank said loan application data for about 25,000 customers leaked, including names, phone numbers, and income.
The breaches were disclosed from September 30. President Lee Jae Myung said on October 6 that signs had emerged of AI agents being deployed in at least some of them, per The Record. Investigators found 33 IP addresses linked to at least 12 countries, and a 28-person police team is on the case. CrowdStrike said the attacker used Artex and is likely a Chinese speaker with a financial motive. No group has been named.
Artex was built as a defensive tool. It automates the security testing companies run against their own systems. It was open source, so anyone could download and modify it. Its developer, who goes by Autumn-27, said on October 8 that the tool had been "abused by some bad actors" and announced it would become closed source, with no further public versions.
That stops new downloads from the original source. It does nothing about copies already taken, as analyst Poe Zhao told AFP. Around 20 Japanese companies have reported similar possible compromises, and police there have not said whether the cases are linked. Lee said AI now makes it possible to hack "with ease even without specialized skills."
Bottom Line
The attacker's advantage was a free tool that runs the tedious steps of an intrusion without a skilled operator. If you hold customer financial data, assume automated credential and vulnerability testing is running against you continuously, and run the same class of tool against yourself first. Closing one tool's source does not withdraw the technique.