AI Agents Used to Compromise Over 440 PaperCut Instances in Widespread Cyberattack

A suspected Russian-speaking cyber actor utilized hundreds of AI agents to exploit recently disclosed vulnerabilities (CVE-2026-81578 and CVE-2026-82078) in PaperCut NG/MF, compromising over 440 instances, primarily in the education sector. This incident demonstrates the increasing sophistication and scale of cyberattacks leveraging AI.