Roundcube CVE-2026-48842: Pre-Auth SQL Injection Actively Exploited in Internet-Facing Mail Servers
Reach 50,000+ security professionals
Sponsor a newsletter slot and put your brand in front of CISOs, engineers, and IT decision-makers.
CVE-2026-48842, a pre-authentication SQL injection flaw in Roundcube Webmail's virtuser_query plugin, is actively exploited against internet-facing mail servers, with the Canadian Centre for Cyber Security confirming attacks on September 21, 2026 -- 117 days after Roundcube shipped the fix in versions 1.6.16 and 1.7.1 on May 24.
The flaw sits in the virtuser_query plugin, which resolves email addresses against a SQL database to find the corresponding virtual user account. PHP's preg_replace() function processes the email address input before the database query. An attacker sends a backslash character before the closing SQL quote in the email field, breaking the quoting logic and allowing arbitrary SQL statements to execute against the Roundcube database. No credentials, no user interaction, and no special access is required. The entire exploit chain fits in a single HTTP POST request.
The consequence of a successful exploit is full read access to the Roundcube database: every email account username, password hash, message body, contact, and active session token stored for every user on that server. If the database includes plaintext passwords or uses weak hashing such as MD5 without a salt, credential recovery is immediate. For organizations where Roundcube passwords match Active Directory, VPN, or cloud SSO credentials, a single mail server compromise hands attackers a lateral movement path across the entire environment.
More than 99,000 Roundcube installations are indexed on Shodan. Organizations running internet-accessible Roundcube on 1.6.x before 1.6.16 or on 1.7.x before 1.7.1 with the virtuser_query plugin configured should treat this as an active incident risk and apply the patch before checking anything else.
How Does CVE-2026-48842 Work?
The CVE-2026-48842 exploit chain begins at the Roundcube login page. When a user submits their email address, the virtuser_query plugin calls PHP's preg_replace() to sanitize the input before passing it to a SQL query that looks up the virtual user entry in the backend database. The vulnerability is a backslash-escape bypass: an attacker sends an email address ending with a backslash character, which escapes the closing single quote that the sanitization function inserts, leaving the SQL string open. Everything after the backslash becomes live SQL, giving the attacker full read control over the query.
A typical exploitation payload looks like user@domain.com' UNION SELECT username,password,3,4,5,6 FROM rc_users -- -. The backslash consumes the quote Roundcube adds, and the UNION SELECT statement appends attacker-controlled column data to the result returned by the virtuser_query lookup. The database executes the injected query, and the server returns the results to the attacker in the same HTTP response -- no guessing, no side channel timing, no error-based inference required.
The technique is a variant of the classic backslash-escape SQL injection documented in OWASP. What makes it dangerous here is the position: the vulnerable code path is pre-authentication, meaning there is no credential or session requirement before the SQL injection payload reaches the database. Any HTTP client with network access to the Roundcube login page can exploit it.
Organizations that have disabled the virtuser_query plugin are not affected. To verify plugin status, check config/config.inc.php for a virtuser_query entry. If the key is present with a non-empty SQL string, the installation is vulnerable.
Target Discovery (T1595)
Attackers use Shodan, Censys, or direct HTTP scanning to enumerate internet-facing Roundcube instances by fingerprinting the login page HTML, version string in source, or response headers.
SQL Injection Probe (T1190)
A crafted POST to /index.php?_task=login submits an email address ending with a backslash. If the virtuser_query plugin is active, the backslash breaks SQL quoting and the injected payload executes. A valid HTTP 200 response with injected column data confirms exploitation.
Credential Database Exfiltration (T1078)
UNION SELECT statements iterate over rc_users, rc_cache, and rc_identities tables to exfiltrate all usernames, password hashes, and email identities. Full database dump completes in seconds over a standard connection.
Credential Cracking and Lateral Movement (T1110.002)
Stolen password hashes are cracked offline. Recovered credentials are tested against email, Active Directory, VPN, and cloud SSO services. Matching passwords provide direct network access without any further exploitation.
Which Organizations Are Being Targeted?
The Canadian Centre for Cyber Security advisory update on September 21 did not attribute the active exploitation to a specific threat actor, but the targeting pattern matches campaigns against government and education mail infrastructure seen across 2025 and 2026. Roundcube is disproportionately deployed in government agencies, universities, self-hosted enterprise environments, and regional hosting providers compared to commercial mail services.
The virtuser_query plugin is nearly universal in non-cloud Roundcube deployments. Organizations running their own mail server with virtual domain routing -- rather than using Microsoft 365 or Google Workspace -- almost always configure virtuser_query to look up email accounts in a database. This makes the vulnerable population broader than simple version checks suggest: any unpatched Roundcube instance with a backend database for virtual user lookup is exploitable.
The attack surface extends to anyone who receives email from or sends email to a compromised organization. Once credential data is extracted from a Roundcube database, attackers pivot to the inbox itself, reading existing email threads to craft targeted follow-on phishing, intercepting financial communications, and accessing embedded two-factor authentication codes sent via email.
Sectors at highest risk include government ministries and state agencies running self-managed email infrastructure, universities and research institutions managing student and faculty webmail, and managed service providers hosting customer Roundcube instances on shared infrastructure. A single compromised hosting provider Roundcube installation can expose thousands of customer email accounts simultaneously.
For comparison, the Cisco Secure Email Gateway CVE-2026-76461 exploitation earlier this month showed how systematically attackers target mail infrastructure: once a mail appliance is compromised, attackers have direct access to every inbound and outbound message the organization sends.
Briefings like this, every morning before 9am.
Threat intel, active CVEs, and campaign alerts, distilled for practitioners. 50,000+ subscribers. No noise.
CVE-2026-48842 Indicators of Compromise
Published IOCs specific to CVE-2026-48842 attacker infrastructure are not yet available. The indicators below are behavioral: specific patterns that appear in logs and filesystem artifacts during and after exploitation of this vulnerability. Hunt for all five in your Roundcube environment before assuming the server is clean.
Subscribe to unlock Indicators of Compromise
Free subscribers unlock full IOC lists, Sigma detection rules, remediation steps, and every daily briefing.
Sigma Detection Rules for CVE-2026-48842
Two generated Sigma rules detect CVE-2026-48842 exploitation across two attack phases: SQL injection attempts at the Roundcube login endpoint and post-exploitation shell spawning from the web process. Both carry experimental status; validate log field mappings against your web server and SIEM configuration before enabling automatic blocking.
Subscribe to unlock Sigma Detection Rules
Free subscribers unlock full IOC lists, Sigma detection rules, remediation steps, and every daily briefing.
Subscribe to unlock Sigma Hunt Queries
Free subscribers unlock full IOC lists, Sigma detection rules, remediation steps, and every daily briefing.
Subscribe to unlock WAF Detection Rules
Free subscribers unlock full IOC lists, Sigma detection rules, remediation steps, and every daily briefing.
How to Patch CVE-2026-48842: Roundcube Remediation Steps
Apply the patch from Roundcube 1.7.1 or 1.6.16 immediately. Both releases contain the complete fix for CVE-2026-48842 and are drop-in replacements for earlier 1.7.x and 1.6.x installations respectively. The update process requires a standard file replacement and database migration command.
Subscribe to unlock Remediation & Mitigation steps
Free subscribers unlock full IOC lists, Sigma detection rules, remediation steps, and every daily briefing.
Why CVE-2026-48842 Matters for Your Security Program
Roundcube deployments are the mail infrastructure blind spot in most enterprise vulnerability programs. Because Roundcube is open-source and free, it is disproportionately deployed in environments with limited dedicated security staffing: government agencies, universities, healthcare systems, and small-to-medium enterprises. The organizations most likely to be running vulnerable versions are also the most likely to have unpatched systems 117 days after the fix was released.
Mail server credential theft has compounding consequences. Email is the identity layer for password resets across most enterprise applications. An attacker who controls a mail database can extract credentials, watch for reset emails in transit, and intercept two-factor authentication codes sent via SMS-to-email gateways. If a compromised Roundcube password matches a domain account, the attacker may pivot from a single mail server compromise to the full Active Directory environment.
The Canadian Centre for Cyber Security's escalation from advisory to active-exploitation warning on September 21 means the window for "we'll get to it next patch cycle" is closed. Every unpatched, internet-accessible Roundcube installation running the virtuser_query plugin should be treated as a potential breach until the update is applied and the credential database is audited.
For organizations that maintain Roundcube as part of a broader mail infrastructure, this is a prompt to review the full mail stack. See the pattern: the Cisco Secure Email Gateway CVE-2026-76461 zero-day from September showed that mail infrastructure components receive fewer proactive security updates than perimeter firewalls, even though a compromised mail server provides similar access to sensitive communications. Mail security deserves the same emergency change procedures applied to VPN and firewall appliances.
The bottom line
CVE-2026-48842 gives any unauthenticated attacker full read access to the Roundcube credential database with a single HTTP request. The Canadian Centre for Cyber Security confirmed active exploitation on September 21, 2026 -- 117 days after the patch. Every Roundcube instance on 1.6.x before 1.6.16 or 1.7.x before 1.7.1 with the virtuser_query plugin enabled is actively at risk. Update to 1.6.16 or 1.7.1, rotate all user credentials from the database, and audit your web server logs for signs of exploitation before the end of today.
This analysis is generic. create a free account to score threats like this against your own stack.
Frequently asked questions
What is CVE-2026-48842 in Roundcube Webmail?
CVE-2026-48842 is a pre-authentication SQL injection vulnerability in the virtuser_query plugin of Roundcube Webmail. An unauthenticated attacker exploits a backslash-escape bypass in PHP's preg_replace() function to inject SQL commands through the email address field on the login page. Successful exploitation grants read access to the full Roundcube database, including email account credentials, message content, and session tokens, without requiring any valid username or password.
Which versions of Roundcube are vulnerable to CVE-2026-48842?
Roundcube versions 1.6.x before 1.6.16 and 1.7.x before 1.7.1 are vulnerable when the virtuser_query plugin is enabled. Roundcube 1.5.x and earlier are not affected by this specific flaw. The virtuser_query plugin is enabled when organizations store email accounts in a relational database instead of system user accounts, which is standard practice in enterprise and hosting deployments using virtual domains.
Is the virtuser_query plugin required for exploitation?
Yes. CVE-2026-48842 only affects Roundcube deployments with the virtuser_query plugin active. This plugin is typically enabled in organizations that route email through virtual user tables in MySQL or PostgreSQL, which is common for shared hosting providers, government email services, and self-hosted enterprise deployments. To check exposure, look for a virtuser_query entry in config/config.inc.php. If that key is present and non-empty, the installation is vulnerable.
Has CVE-2026-48842 been exploited in real attacks?
Yes. The Canadian Centre for Cyber Security updated advisory AV26-503 on September 21, 2026 to confirm active exploitation of CVE-2026-48842 in the wild. The flaw was patched in May 2026, giving defenders 117 days before exploitation began. Attackers are specifically targeting organizations that have not yet applied the 1.6.16 or 1.7.1 security updates, with internet-facing Roundcube deployments the primary target.
How do I patch CVE-2026-48842 on my Roundcube server?
Update Roundcube to version 1.6.16 from any 1.6.x release or to 1.7.1 from any 1.7.x release. Both versions fix CVE-2026-48842. If patching immediately is not possible, disable the virtuser_query plugin by removing or commenting out the virtuser_query line in config/config.inc.php and restarting the web server. This removes the vulnerable code path without disabling core Roundcube functionality. Resume full use only after updating to a fixed version.
What data can an attacker steal by exploiting CVE-2026-48842?
Successful exploitation gives an attacker read access to the Roundcube database. This typically includes email account usernames and password hashes, email message content, contact lists, and active session tokens. If the database uses weak hashing or plaintext storage, credential recovery is immediate. The stolen credentials then enable account takeover, password spray against other services using the same passwords, and network pivot if Roundcube credentials match Active Directory or VPN accounts.
How do I detect if my Roundcube server was hit by CVE-2026-48842?
Check web server access logs for POST requests to index.php containing the _task=login parameter with backslash characters, single quotes, or SQL keywords in the submitted email field. Review your database audit or slow query logs for unexpected SELECT statements referencing information_schema tables or using UNION clauses executed by the Roundcube database user. Look for authentication log entries showing failed logins from an IP that later appears as successful, which can indicate SQL injection bypass.
Does exploiting CVE-2026-48842 require user interaction?
No. CVE-2026-48842 is a pre-authentication vulnerability requiring no valid credentials and no action from any user. An attacker sends one crafted HTTP POST request to the Roundcube login page. No email needs to be opened. No user account is needed. This means automated scanning and exploitation is straightforward, which is why the Canadian Centre for Cyber Security escalated from a standard advisory to active-exploitation warning 117 days after the patch was available.
Sources & references
- Canadian Centre for Cyber Security Advisory AV26-503 (September 21, 2026)
- Roundcube Security Release 1.7.1 (May 24, 2026)
- The Hacker News: Roundcube Pre-Auth SQL Injection Flaw Actively Exploited
- TechRepublic: Attackers Target Unpatched Roundcube Servers With CVE-2026-48842
- MITRE ATT&CK: T1190 Exploit Public-Facing Application
Free resources
Critical CVE Reference Card 2025–2026
25 actively exploited vulnerabilities with CVSS scores, exploit status, and patch availability. Print it, pin it, share it with your SOC team.
Ransomware Incident Response Playbook
Step-by-step 24-hour IR checklist covering detection, containment, eradication, and recovery. Built for SOC teams, IR leads, and CISOs.
Daily briefings + Google Top Stories in one step.
Enter your email and we'll subscribe you to the daily briefing and add Decryption Digest to your Google Search feed at the same time.

Founder & Cybersecurity Evangelist, Decryption Digest
Cybersecurity professional with expertise in threat intelligence, vulnerability research, and enterprise security. Covers zero-days, ransomware, and nation-state operations for 50,000+ security professionals every morning.
