Introduction: PaperCut AI Attack — Why It Matters
The PaperCut AI Attack reportedly involved hundreds of autonomous AI agents exploiting vulnerabilities in PaperCut NG/MF, potentially compromising at least 440 servers across 395 organizations in 48 countries. The campaign is reportedly linked to a Russian-speaking threat actor and highlights how AI-assisted automation can dramatically accelerate cyber intrusions.
According to the reported PaperCut AI Attack findings, attackers combined AI agents with established offensive-security tools to move from initial access to credential theft and domain-level compromise in minutes. PaperCut has separately confirmed active exploitation of its NG/MF products and published emergency patches for the two vulnerabilities involved.
What Is PaperCut NG/MF?
PaperCut NG and PaperCut MF are print-management platforms used by organizations to manage printing, user access, accounting and related workflows.
PaperCut’s current security advisory identifies CVE-2026-81578, an authentication-bypass vulnerability, and CVE-2026-82078, an unsafe dynamic class-loading vulnerability. PaperCut rates the latter as critical with a CVSS score of 9.4 and the authentication bypass as high with a score of 8.8.
Organizations running internet-accessible PaperCut Application Servers therefore face particular risk if emergency patches and access restrictions have not been applied.
Who Is Behind the Attack?
The PaperCut AI Attack has reportedly been attributed to a Russian-speaking threat actor, although the actor’s ultimate objective remains unclear.
The reported operation allegedly used hundreds of autonomous AI agents, with an OpenAI Codex harness and a DeepSeek model helping automate portions of reconnaissance, exploitation and post-compromise activity.
The attackers reportedly also used established tools including:
- Mimikatz
- Certipy
- Rubeus
- Impacket
This combination suggests that the AI was not necessarily replacing conventional offensive tooling. Instead, it reportedly helped coordinate and accelerate the use of existing tools.
PaperCut AI Attack: Technical Breakdown
Timeline of Events
The reported PaperCut AI Attack demonstrates how quickly automated attack chains can progress.
According to the reported findings:
- AI agents allegedly compromised 11 organizations in only 26 seconds.
- One U.S. high school network reportedly progressed from initial access to Domain Administrator in about seven minutes.
- Attackers allegedly harvested credentials from LSASS and the Windows Registry.
- The campaign reportedly exploited the noPac vulnerabilities, CVE-2021-42278 and CVE-2021-42287.
- Rogue Domain Administrator accounts were reportedly created during successful intrusions.
- Some compromised environments allegedly experienced DCSync activity, allowing attackers to obtain credentials from the NTDS.DIT database.
The speed of these operations is significant because defenders may have only a narrow window to detect and contain an intrusion once an exposed server is compromised.
What Systems Were Affected?
Reportedly affected systems included PaperCut servers and, in successful attack chains, connected Windows environments.
The reported victim breakdown includes:
- 440 compromised systems across 395 organizations
- Victims in 48 countries
- 98 victims in the United States
- 204 compromised systems associated with educational institutions
- Domain environments where attackers allegedly obtained elevated privileges
These figures remain attributed to the reported research and should not be interpreted as a complete official victim list.
Potential Risks & Impact
Credential and Identity Risk
The PaperCut AI Attack can allow attackers to move beyond the initially compromised PaperCut server. DCSync activity is particularly serious because it can expose password hashes and other authentication material stored within Active Directory.
Business and Reputational Risk
A compromised print-management server can become an entry point into a broader enterprise network. Once attackers obtain privileged credentials, they may be able to access additional systems, disrupt operations or establish persistent access.
Regulatory and Compliance Risk
Organizations handling sensitive employee, student or customer information may face additional investigation and notification requirements if the intrusion results in unauthorized data access.
For organizations tracking major incidents and regulatory developments, CyberNexora’s cyber incidents coverage provides additional reporting.
Official Response / Statement
PaperCut has confirmed that it is investigating active exploitation affecting PaperCut NG/MF and has issued emergency patches. The company recommends restricting internet access to PaperCut Application Servers and installing its latest emergency release.
PaperCut’s official security advisory
GreyNoise is reportedly coordinating with incident-response partners to notify affected organizations and publishing fresh indicators of compromise through its public repositories.
Industry Context: Why AI-Assisted Attacks Are Increasing
The reported PaperCut AI Attack illustrates a broader shift in offensive cybersecurity: attackers can use AI to automate repetitive tasks that previously required significant human involvement.
Instead of manually conducting reconnaissance, testing credentials and selecting post-exploitation actions, autonomous agents can reportedly execute multiple steps rapidly. The most important development is therefore not simply the use of an AI model, but the ability to connect that model with real security tools and allow it to operate at scale.
The reported instruction failures are also notable. Agents were allegedly told to avoid 28 countries, yet victims were still recorded in some of those regions. That suggests autonomous systems may not always follow operational constraints reliably.
Organizations can review CyberNexora’s Learn & Protect resources for broader defensive guidance.
How to Protect Your Organization
- Patch PaperCut immediately: Install the latest applicable PaperCut NG/MF emergency release and follow the vendor’s upgrade guidance.
- Restrict internet exposure: Do not leave PaperCut Application Server web interfaces openly accessible from untrusted networks.
- Review authentication logs: Investigate unusual administrative activity and unexpected account creation.
- Monitor privileged accounts: Look for suspicious Domain Admin additions, privilege escalation and abnormal authentication patterns.
- Hunt for credential theft: Investigate suspicious LSASS access, credential-dumping activity and unexpected DCSync operations.
- Review PaperCut logs: Look for missing, truncated or unexpectedly deleted server logs and suspicious activity involving the application server.
- Segment critical systems: Limit communication between print-management infrastructure and sensitive Active Directory resources.
- Use layered detection: Combine endpoint, identity, network and application-level monitoring rather than relying on a single security control.
PaperCut’s advisory specifically recommends immediately restricting public access to internet-facing Application Servers.
Indicators of Compromise (IoCs)
Organizations investigating potential exposure should look for:
- Suspicious activity involving PaperCut Application Servers
- Unexpected or deleted
server.logfiles - LSASS or Registry credential-harvesting activity
- noPac exploitation indicators
- Unauthorized Domain Administrator accounts
- Suspicious DCSync operations
- Unexpected use of Mimikatz, Certipy, Rubeus or Impacket
- Abnormal post-exploitation activity originating from PaperCut infrastructure
Key Takeaways
- The reported PaperCut AI Attack involved hundreds of autonomous AI agents.
- At least 440 servers across 395 organizations were reportedly compromised.
- Education was reportedly the largest affected sector, accounting for 204 systems.
- AI-assisted automation allegedly accelerated attacks from initial access to privileged compromise.
- PaperCut has confirmed active exploitation and released emergency patches for the affected vulnerabilities.
Conclusion: PaperCut AI Attack and What Happens Next
The PaperCut AI Attack demonstrates the potential speed and scale of autonomous AI-assisted cyber operations. While conventional tools such as credential dumpers and Active Directory attack utilities remain central to the intrusion chain, AI can potentially coordinate those capabilities far faster than a human operator.
Organizations responding to the PaperCut AI Attack should prioritize patching, restrict public exposure and investigate systems for signs of credential theft or privilege escalation. Further attribution, victim notifications and the attacker’s ultimate objective remain areas to watch as investigations continue.
For more incident coverage, readers can follow CyberNexora’s Cyber Incidents category.
Frequently Asked Questions (FAQs)
The PaperCut AI Attack refers to a reported campaign in which autonomous AI agents allegedly helped exploit PaperCut NG/MF vulnerabilities. The reported campaign affected hundreds of organizations worldwide.
At least 440 servers across 395 organizations in 48 countries were reportedly compromised. The figures come from the reported campaign findings and may change as investigations continue.
The reported campaign targeted CVE-2026-81578 and CVE-2026-82078. PaperCut identifies the first as an authentication bypass and the second as an unsafe dynamic class-loading vulnerability.
The activity has reportedly been associated with a Russian-speaking threat actor. The actor’s ultimate objective has not been conclusively established.
Organizations should install the latest emergency patches, restrict public access to PaperCut Application Servers, monitor privileged accounts and investigate suspicious credential or Active Directory activity. PaperCut recommends immediate access restrictions for internet-facing servers.
A ransomware connection has not been established from the supplied findings. It remains unclear whether the compromised access will be sold, used for extortion or pursued for another objective.