Google Threat Intelligence Group and Mandiant identified a system called exploit_agent that automates vulnerability detection, malicious code generation, and attack execution with minimal human involvement. The system analyzed software updates to find recently patched vulnerabilities, then targeted websites that had not yet installed the corresponding security fixes. The investigation also uncovered stolen data, network reconnaissance scripts, and webshells that provided remote access to compromised servers. Separately, suspected North Korean hackers abused AI-powered development tools to infiltrate cryptocurrency organizations in North America and Europe, modifying Claude Code configurations to deploy a backdoor.
Source: Read the original article


