A threat actor — likely Russian-speaking — used hundreds of AI agents to design, test, and launch a mass-exploitation campaign against PaperCut print management servers, compromising at least 440 instances tied to 395 distinct organizations in 48 countries. The campaign began August 31 and was first documented by threat intelligence firm GreyNoise.
The attackers combined OpenAI’s Codex and DeepSeek language models with standard offensive tools to build working exploits for two PaperCut NG/MF vulnerabilities: CVE-2026-81578 and CVE-2026-82078. The AI handled everything from identifying targets via the Netlas internet-scanning platform to refining exploit code and executing post-exploitation steps. GreyNoise reports the attacker went from an empty workspace to first remote code execution against a real victim in under four hours — and once the campaign launched at scale, compromised at least 11 organizations in 26 seconds. In one case, a high school was fully owned (initial access to domain administrator) in seven minutes.
The education sector accounted for roughly half of all breaches. Attackers harvested credentials from 280 victims, obtained OS or domain secrets from 147, and gained administrator privileges at 12 organizations. After gaining a foothold, the attacker used LSASS memory dumps, the noPac technique (targeting CVE-2021-42278 and CVE-2021-42287), and DCSync to extract a complete dump of domain credentials. The toolkit included Mimikatz, BloodHound, Rubeus, Certipy, Impacket, Ligolo-ng, and custom Rust-based credential collectors.
How to check if you’re affected
Affected versions are PaperCut NG and MF servers that have not yet applied the security patches for CVE-2026-81578 and CVE-2026-82078. Log into your PaperCut admin console and check the version under Help → About. If your server is internet-accessible and unpatched, assume active scanning is already underway and patch immediately or take the server offline.