PaperCut Bug Mayhem: Print Server Leads Hackers to Domain Control

A critical vulnerability in PaperCut MF was exploited to turn a simple print server into a launching pad for a full domain controller takeover. Attackers targeted version 24.0.2, build 69746, using two zero-day flaws—CVE-2026-81578 and CVE-2026-82078—that allowed them to change server settings without authentication and execute Java bytecode in the server’s security context. The compromised server, accessible from the internet, gave the intruders a path to Active Directory.

The breach was discovered on August 31, 2026, at an education-sector client. Within less than 48 hours, the attackers had escalated access from the print server to a domain controller.

How the Attack Unfolded

Initial compromise came through exploiting those two PaperCut MF zero-day vulnerabilities. Once inside, the attackers used the card or ID lookup field to inject Java code. This led to the installation of a memory-based loader and a web shell—used to execute commands, access configuration data, clean logs and application databases, and block other exploits. The web shell also intercepted certain server requests, effectively locking out unrelated attempts to use the same flaws.

From that foothold, the attackers deployed an implant dubbed AdaptixC2, which was hidden inside a tampered Microsoft Copilot binary. The installer lay dormant for about one day before full activity unfolded.

Privilege Escalation & Domain Controller Capture

Once the malicious implant was active, reconnaissance began. The attackers cataloged networks, domain trusts, admin accounts, and more. They identified a process running with domain-privileged service account rights, stole its access token, then relaunched the implant under those elevated credentials. No password was stolen directly—access token abuse sufficed.

The next move: copying payloads to the domain controller via shared network storage, tweaking the Windows Plug and Play service to launch the payload stealthily, and reverting the service to its original path to avoid suspicion. Once on the domain controller, credentials were dumped from memory and the registry. They enabled Windows Restricted Admin mode and used a compromised NTLM hash to authenticate over Remote Desktop Protocol, ultimately extracting a copy of the Active Directory database—a vault containing every domain account’s password hash.

The attackers also armored their tools with obfuscation: encrypted configuration data, scrambled code logic, and environment checks meant many public sandbox tools failed to analyze the malware properly.

Mitigation & Lessons Learned

As an urgent countermeasure, it’s vital for organizations using PaperCut MF or NG to update to the most recent version. Servers exposed to the internet should restrict access to trusted IP ranges only. Watch for red flags like missing or truncated logs, unexpected service reconfigurations, unusual child processes spawned by the PaperCut service, and signs of post-exploit malicious behavior.

Minimizing the privileges assigned to service accounts and ensuring continuous monitoring of endpoints are also key defenses. Analysts recommend following vendor-supplied security advisories, reviewing indicators such as the ones documented in this case, and isolating affected hosts while remediating.

Indicators of Compromise (IoCs) include download URLs for the AdaptixC2 implant, multiple file hashes for the trojanized binaries (Microsoft Copilot, PulseSecure.exe), and variants of the loader and web shell for the Java bytecode’s Jakarta and Javax payloads.

This incident underscores how peripheral infrastructure—print servers—can become critical attack vectors when left exposed. It’s a stark reminder that zero-day vulnerabilities are a matter of when, not if. Administrators must prioritize patching, exposure-hardening, and vigilant monitoring to defend identity-sensitive systems against increasingly sophisticated threats.