Progress has disclosed 31 vulnerabilities in MarkLogic Server that could lead to privilege escalation, server-side code execution, authentication bypass, credential exposure, and data compromise.
In its security bulletin dated October 12, 2026, the company urges customers to install updates, warning that affected instances may operate normally without any prior signs of exploitation.
The bulletin covers MarkLogic Server versions 12.0 through 12.0.3 and 11.0 through 11.3.6. The impact of each vulnerability varies depending on the platform, enabled features, and configuration.
Progress identifies MarkLogic Server versions 12.1 and 11.3.7 as the patched releases. Progress has not fully assessed retired versions, including version 10 and earlier, and will not provide fixes.
Several flaws compromise authorization boundaries. CVE-2026-85947 allows attackers with forest-management privileges to create a persistent, unauthorized administrator account.
CVE-2026-85948 reveals an XQuery-injection vulnerability in a deprecated Management API package path, potentially granting attackers with manage-user-level privileges persistent administrative control.
CVE-2026-85955 enables remote attackers with query-execution access to invoke an XQuery evaluation function without the required additional privilege, which may lead to executing server-side code with elevated permissions.
Similarly, CVE-2026-85954 evaluates attacker-controlled values submitted to certain higher-order XQuery functions in a privileged security context.
The Query Console introduces another execution vector. CVE-2026-85960 involves predictable cross-site request forgery (CSRF) tokens, allowing an attacker to lure an authenticated user to a malicious page. This could let an attacker submit unauthorized requests and execute code under that user’s permissions.
These scenarios require specific prerequisites; the bulletin does not classify every code-execution issue as susceptible to unauthenticated remote exploitation.
CVE-2026-85962 affects SAML authentication and could let unauthenticated remote attackers submit assertions that should be rejected. CVE-2026-85958 addresses the acceptance of invalid authentication tokens, while CVE-2026-9191 inadequately isolates evaluated user code from HTTP request, response, and session states.
Cloud deployments are susceptible to server-side request forgery through CVE-2026-85951 on Azure and CVE-2026-85952 on affected Linux installations. Authenticated attackers could access internal cloud services and potentially obtain credentials or tokens.
Additional vulnerabilities include XML external entity processing issues, arbitrary server-file disclosure, sensitive information leaks in logs, and insufficient auditing of authentication.
Memory-related defects affect JavaScript integration and can lead to use-after-free errors, out-of-bounds reads, and out-of-bounds writes. The disclosed consequences include instability, memory leakage, and limited data modification capabilities.
Progress recommends that affected customers upgrade their deployments to versions 12.1 or 11.3.7. Users of retired versions should migrate to a supported, fixed release. Progress has already patched Data Cloud deployments.
Administrators should review the compatibility documentation before deployment, since security changes can affect APIs, privilege requirements, authentication, logging, XML parsing, and transport behavior. Progress’s documentation explicitly identifies product changes that could lead to application compatibility issues.
Older MarkLogic Content Pump clients must match the fixed server version. Custom RPM-based Linux installations also require curl for the protection of shipped cloud metadata.
Existing log permissions necessitate separate remediation. The lack of audit evidence does not rule out exploitation, so the appearance of normal operation is an unreliable assurance against vulnerabilities.
All CVE Details
| CVE | CWE / Vulnerability Type |
|---|---|
| CVE-2026-9191 | CWE-346 — Origin Validation Error |
| CVE-2026-85947 | CWE-269 — Improper Privilege Management |
| CVE-2026-85948 | CWE-269 — Improper Privilege Management |
| CVE-2026-85949 | CWE-532 — Sensitive Information in Log Files |
| CVE-2026-85950 | CWE-269 — Improper Privilege Management |
| CVE-2026-85951 | CWE-918 — Server-Side Request Forgery |
| CVE-2026-85952 | CWE-918 — Server-Side Request Forgery |
| CVE-2026-85953 | CWE-22 / CWE-269 — Path Traversal and Improper Privilege Management |
| CVE-2026-85954 | CWE-269 — Improper Privilege Management |
| CVE-2026-85955 | CWE-269 — Improper Privilege Management |
| CVE-2026-85956 | CWE-269 — Improper Privilege Management |
| CVE-2026-85957 | CWE-611 — XML External Entity Processing |
| CVE-2026-85958 | CWE-287 — Improper Authentication |
| CVE-2026-85959 | CWE-778 — Insufficient Logging |
| CVE-2026-85960 | CWE-352 — Cross-Site Request Forgery |
| CVE-2026-85961 | CWE-532 — Sensitive Information in Log Files |
| CVE-2026-85962 | CWE-287 — Improper Authentication |
| CVE-2026-85963 | CWE-918 / CWE-321 — SSRF and Hard-Coded Cryptographic Key |
| CVE-2026-85964 | CWE-22 — Path Traversal |
| CVE-2026-85965 | CWE-416 — Use After Free |
| CVE-2026-85966 | CWE-200 — Exposure of Sensitive Information |
| CVE-2026-85967 | CWE-79 — Cross-Site Scripting |
| CVE-2026-85968 | CWE-125 — Out-of-Bounds Read |
| CVE-2026-85969 | CWE-787 — Out-of-Bounds Write |
| CVE-2026-85970 | CWE-611 — XML External Entity Processing |
| CVE-2026-85971 | CWE-732 — Incorrect Permission Assignment |
| CVE-2026-85972 | CWE-303 — Incorrect Authentication Algorithm Implementation |
| CVE-2026-85973 | CWE-319 — Cleartext Transmission of Sensitive Information |
| CVE-2026-85974 | CWE-297 — Improper TLS Certificate Hostname Validation |
| CVE-2026-85975 | CWE-319 — Cleartext Transmission of Sensitive Information |
| CVE-2026-85976 | CWE-113 — Improper Neutralization of CRLF Sequences in HTTP Headers |
Update defenses against new malware as soon as it appears. Integrate TI Feeds with data from 16K SOCs.
Attackers abused compromised ConnectWise Cloud Backup credentials to access Microsoft 365 customer environments, prompting ConnectWise…
Anthropic says an internal Claude research model reconstructed how the National Security Agency generated seeds…
Two WordPress Core vulnerabilities can be chained to execute remote code without authentication, according to…
A newly disclosed OpenBSD kernel exploitation chain combines a shared memory accounting flaw with a…
GhostAction attackers have expanded their GitHub credential theft campaign with a second stage that disguises…
Indian authorities have warned that SideCopy hackers are targeting Indian organizations, including academic institutions, with…