Skip to main content

MongoDB Server EUVDEUVD-2026-56908

| CVE-2026-18705 HIGH
Reliance on Untrusted Inputs in a Security Decision (CWE-807)
2026-08-11 mongodb GHSA-f96f-54rq-c385
7.1
CVSS 4.0 · Vendor: mongodb
Share

Severity by source

Vendor (mongodb) PRIMARY
7.1 HIGH
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X
vuln.today AI
6.5 MEDIUM

Network-accessible with low complexity; authenticated low-privilege access required (PR:L); confidentiality-only impact with no scope change beyond the vulnerable collection.

3.1 AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:N
4.0 AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:N/VA:N/SC:N/SI:N/SA:N

Primary rating from Vendor (mongodb).

CVSS VectorVendor: mongodb

Attack Vector
Network
Attack Complexity
Low
Privileges Required
Low
User Interaction
None
Scope
X

Lifecycle Timeline

3
Patch available
Aug 11, 2026 - 20:19 EUVD
Analysis Generated
Aug 11, 2026 - 20:14 vuln.today
CVE Published
Aug 11, 2026 - 18:37 cve.org
HIGH 7.1

DescriptionCVE.org

An issue in MongoDB Server's Atlas Vector Search feature could allow an authenticated user with read access to one view to retrieve documents from a different, protected view over the same underlying collection. This is due to insufficient handling of certain user-supplied fields when constructing an internal request forwarded to the search process.

AnalysisAI

Atlas Vector Search in MongoDB Server exposes a horizontal authorization bypass allowing an authenticated low-privileged user to read documents from a view they are not authorized to access, crossing view-level access control boundaries within the same underlying collection. The flaw arises from MongoDB Server insufficiently validating user-supplied fields before forwarding internal requests to the search process, enabling the search backend to return documents outside the scope of the caller's authorized view. No public exploit code or CISA KEV listing has been identified at time of analysis.

Technical ContextAI

MongoDB Server implements view-level access control as a mechanism for row-level security, allowing administrators to expose filtered subsets of a collection's documents to different users or roles. Atlas Vector Search is a semantic similarity query feature that enables embedding-based retrieval over collections. CWE-807 (Reliance on Untrusted Inputs in a Security Decision) is the root cause class: the server uses attacker-controlled field values from the incoming query to construct an internal forwarded request to the search process, and that forwarded request fails to preserve or enforce the originating view's filter boundary. As a result, the search process operates against the underlying collection rather than the restricted view, leaking documents the caller was never authorized to see. The CPE cpe:2.3:a:mongodb:mongodb_server:*:*:*:*:*:*:*:* indicates MongoDB reported this across the server product line without specifying a bounded version range in available data.

RemediationAI

Monitor the MongoDB JIRA ticket at https://jira.mongodb.org/browse/SERVER-129618 for a vendor-released patch and apply it immediately upon availability; no exact fixed version has been confirmed in available data. As a compensating control prior to patching, administrators should audit role-based access controls and temporarily revoke Atlas Vector Search access for users who do not have legitimate business need for cross-view query capabilities, using MongoDB's built-in privilege system to restrict the relevant search roles or actions. Additionally, consider enforcing network-level restrictions to limit which authenticated clients can reach the Atlas Vector Search endpoint, reducing exposure to lower-trust user tiers. Note that revoking Vector Search access will disrupt semantic similarity workloads for affected users, so scope restrictions carefully. Enable MongoDB audit logging on search operations to detect anomalous cross-view access patterns during the remediation window.

CVE-2013-1892 MEDIUM POC
6.0 Oct 01

MongoDB before 2.0.9 and 2.2.x before 2.2.4 does not properly validate requests to the nativeHelper function in SpiderMo

CVE-2013-3969 MEDIUM POC
6.5 Oct 01

The find prototype in scripting/engine_v8.h in MongoDB 2.4.0 through 2.4.4 allows remote authenticated users to cause a

CVE-2018-13863 HIGH POC
7.5 Jul 10

The MongoDB bson JavaScript module (also known as js-bson) versions 0.5.0 to 1.0.x before 1.0.5 is vulnerable to a Regul

CVE-2026-25887 HIGH POC
7.2 Mar 06

Remote code execution in Chartbrew versions prior to 4.8.1 allows authenticated attackers with high privileges to execut

CVE-2019-2386 HIGH POC
7.1 Aug 06

After user deletion in MongoDB Server the improper invalidation of authorization sessions allows an authenticated user's

CVE-2012-6619 MEDIUM POC
6.4 Mar 06

The default configuration for MongoDB before 2.3.2 does not validate objects, which allows remote authenticated users to

CVE-2026-3431 CRITICAL
9.8 Mar 02

SimStudio below 0.5.74 has a missing authorization on MongoDB tool endpoints that allows attackers to execute arbitrary

CVE-2024-8654 CRITICAL
9.8 Sep 10

MongoDB Server may access non-initialized region of memory leading to unexpected behaviour when zero arguments are calle

CVE-2024-6376 CRITICAL
9.8 Jul 01

MongoDB Compass may be susceptible to code injection due to insufficient sandbox protection settings with the usage of e

CVE-2024-1351 CRITICAL
9.8 Mar 07

Under certain configurations of --tlsCAFile and tls.CAFile, MongoDB Server may skip peer certificate validation which ma

CVE-2020-7610 CRITICAL
9.8 Mar 30

All versions of bson before 1.1.4 are vulnerable to Deserialization of Untrusted Data. Rated critical severity (CVSS 9.8

CVE-2026-19001 CRITICAL
9.5 Aug 12

Memory corruption in the MongoDB BI Connector ODBC Driver (fixed in v1.4.9) occurs when an oversized catalog, schema, or

Share

EUVD-2026-56908 vulnerability details – vuln.today

This site uses cookies essential for authentication and security. No tracking or analytics cookies are used. Privacy Policy