Hackers use big language models and AI tools to attack South Korean financial institutions, resulting in data breaches for 68000 people

The CrowdStrike report shows that hackers used big language models and AI penetration tool ARTEX to attack multiple financial institutions in South Korea from late September to early October. Hackers used a dual server architecture and orchestrated attacks using Claude Code, GLM-5.3, Grok 4.6, and DeepSeek v4.1-flash, resulting in the leakage of personal data of approximately 68000 people from at least seven institutions including KB National Bank, Shinhan Bank, and Hana Bank. (Source: CrowdStrike)