Posted in

China-Linked Hackers Exploit Critical Roundcube Flaws to Target US and Canadian Universities

A newly discovered cyber espionage campaign has revealed that a suspected China-aligned threat group is targeting universities across the United States and Canada by exploiting critical vulnerabilities in the Roundcube webmail platform.

According to cybersecurity researchers at Proofpoint, the attackers are primarily focusing on physics and engineering departments, especially those involved in astrophysics, particle physics, or research connected to national security.

Critical Roundcube Flaws Used to Gain Access

The attackers are taking advantage of multiple high-severity vulnerabilities in Roundcube, including:

  • CVE-2024-42009 (CVSS 9.3) – A cross-site scripting (XSS) vulnerability.
  • CVE-2025-49113 (CVSS 9.9) – A post-authentication remote code execution (RCE) vulnerability.

Although these vulnerabilities have already been patched, systems running outdated versions of Roundcube remain vulnerable to exploitation.

Proofpoint has assigned the threat cluster the name UNK_MassTraction, with the campaign first observed in May 2026.

Universities and Researchers Are the Primary Targets

The campaign specifically targets administrators and professors working in departments conducting advanced scientific research or projects with national security implications.

Researchers found that attackers distributed phishing emails using compromised email accounts and domains with weak DMARC policies, allowing them to spoof legitimate senders more effectively.

Interestingly, the phishing emails themselves used generic themes, suggesting the campaign may be broader than the currently identified victims.

Opening an Email Is Enough to Trigger the Attack

One of the most concerning aspects of the campaign is how little interaction is required from victims.

Simply opening a malicious email within the vulnerable Roundcube webmail interface can trigger the XSS exploit. The exploit executes malicious JavaScript directly inside the victim’s browser, giving attackers access to the mail server.

Proofpoint believes the attackers conducted detailed reconnaissance beforehand to identify organizations still running vulnerable Roundcube versions before launching the phishing campaign.

IceCube Malware Steals Credentials and Session Data

After successfully exploiting the XSS vulnerability, the attackers deploy a JavaScript payload known as IceCube.

The malware is designed to collect valuable information, including:

  • Browser-stored credentials
  • Authentication cookies
  • Two-factor authentication (2FA) data
  • Browser language settings
  • Screen resolution
  • Form input values

The stolen data is then transmitted to an attacker-controlled server using HTTP POST requests.

Second Exploit Provides Persistent Server Access

IceCube doesn’t stop at credential theft.

Using the victim’s active session and CSRF token, it exploits the second Roundcube vulnerability (CVE-2025-49113) to execute code on the mail server itself.

Once successful, attackers deploy one of two persistence mechanisms:

  • SquareShell, a PHP-based web shell
  • VShell, a sophisticated remote administration tool

SquareShell enables attackers to execute arbitrary commands through the Roundcube endpoint:

plugins/newmail_notifier/mail_preview.php

If the web shell installation fails, the attackers automatically switch to an alternative infection method that ultimately installs VShell.

Attack Chain Evolved in June 2026

Researchers noted that the campaign became more advanced in June 2026.

Previously, failed web shell deployments caused the attack to terminate. The updated version now executes a shell script that downloads an ELF loader known as SNOWLIGHT, which installs VShell based on the victim’s operating system architecture.

SNOWLIGHT has previously been associated with Chinese cyber-espionage operations, including activity linked to the threat cluster UNC5174.

This overlap suggests that several China-linked hacking groups may be sharing specialized malware and exploitation tools within private ecosystems.

Deferred Triggers Help Attackers Stay Hidden

IceCube also includes sophisticated persistence techniques designed to maximize the chances of successful exploitation.

The malware monitors user behavior by detecting events such as:

  • Closing the browser tab
  • Switching tabs
  • Moving the mouse outside the browser window
  • Clicking the logout button

Whenever one of these actions occurs, IceCube attempts to re-exploit the Roundcube vulnerability and notifies its command-and-control (C2) server that the victim has left the session.

After completing its objectives, the malware deliberately logs users out and destroys both legitimate and malicious sessions, reducing forensic evidence and making incident investigation more difficult.

VShell Provides Long-Term Remote Control

Written in the Go programming language, VShell functions as a full-featured remote administration tool, offering capabilities similar to Cobalt Strike.

Once deployed, it enables attackers to:

  • Execute remote commands
  • Maintain persistent access
  • Move laterally within networks
  • Conduct additional reconnaissance
  • Deploy follow-on malware

VShell has appeared in several previous campaigns attributed to China-aligned threat actors.

A Shift in Nation-State Targeting

Historically, Roundcube vulnerabilities have been heavily exploited by Russian state-sponsored groups.

However, Proofpoint says this campaign represents the first publicly documented case of a China-linked threat actor systematically exploiting Roundcube vulnerabilities during espionage operations.

The researchers also emphasized that attackers increasingly view email servers as high-value entry points rather than just communication platforms.

Security Recommendations

Organizations using Roundcube should immediately review their email infrastructure and apply all available security updates.

Recommended defensive measures include:

  • Update Roundcube to the latest patched version.
  • Enable strong DMARC, SPF, and DKIM policies.
  • Monitor webmail servers for unusual activity.
  • Watch for unauthorized web shells or persistence mechanisms.
  • Enforce multi-factor authentication across administrative accounts.
  • Continuously monitor for indicators of compromise.

Final Thoughts

The UNK_MassTraction campaign demonstrates how sophisticated threat actors continue to evolve their techniques by combining phishing, zero-day and n-day vulnerabilities, credential theft, and stealthy persistence mechanisms.

As email servers increasingly become gateways into enterprise networks, organizations must treat them with the same level of protection as VPN gateways, remote access infrastructure, and other internet-facing systems. Prompt patching, continuous monitoring, and layered security controls remain essential for defending against advanced cyber espionage campaigns.

Leave a Reply

Your email address will not be published. Required fields are marked *