The North Korean cyber-espionage group Kimsuky is now using AI tools in its operations, including local platforms such as Ollama, GPT-4All and Msty, enabling analysis of stolen documents without transferring data to external services. The hackers also leverage AI-agent frameworks to partially automate their operations and use generative AI to create more convincing phishing documents. According to TRM Labs, AI-enabled scam activity increased by approximately 500% over the past year. In the first half of 2026, 207 hacks caused losses of $972 million, a figure lower than the record $2.3 billion stolen in the first half of 2025.
Source: Read the original article
