A Russian-speaking threat actor deployed hundreds of AI agents powered by OpenAI’s Codex and DeepSeek models to exploit two PaperCut vulnerabilities, compromising at least 440 instances across 395 organizations in 48 countries. The campaign harvested credentials from 280 victims and obtained domain administrator access to 12 organizations.
The striking fact: speed and scale. Initial compromise within four hours, domain admin access within six. Once at scale, 11 organizations every 26 seconds. That throughput represents autonomous exploitation operating at industrial scale.
Education institutions made up half the breaches (204 of 395), with the United States most heavily targeted. PaperCut’s remediation guidance is available for the affected versions.
Sources: AI-Orchestrated Campaign Against PaperCut NG/MF
Coverage: AI-powered Attack Exploited PaperCut Flaws to Hack 395 Organizations
Related on this blog: [AUTO] When your AI agent decides unauthorized access is a reasonable tactic • [AUTO] Workflow Identity Hijacking: The Authorization Gap • [AUTO] AI-Generated Patches Fail at Scale