What is the CVSS score of EUVD-2026-58289?

EUVD-2026-58289 has a CVSS 3.1 base score of 7.1 (High). CVSS vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:L. EPSS exploitation probability: 0.3%.

Which versions of Kibana are affected by EUVD-2026-58289?

Kibana contains a cross-space authorization bypass allowing authenticated users to access machine learning model outputs and modify deployments across unauthorized spaces, creating data exposure and…

How to fix or mitigate EUVD-2026-58289?

Within 24 hours, identify all Kibana instances in production and verify current versions against the 8.x, 9.4, and 9.5 release lineages. Within 7 days, upgrade affected Kibana deployments to patched versions 8.19.20, 9.4.5, or 9.5.1 depending on your lineage, prioritizing instances that host machine learning models. Within 30 days, verify patch deployment completion across your entire environment, review access logs for unauthorized cross-space model inference or deployment modifications, and document any suspicious access patterns.

Kibana EUVD-2026-58289

| CVE-2026-72629 HIGH

Authorization Bypass Through User-Controlled Key (CWE-639)

2026-08-13 elastic

GHSA-397q-qwg5-f6mc

Authentication Bypass Elastic Kibana

7.1

CVSS 3.1 · Vendor: elastic

Severity by source

Vendor (elastic) PRIMARY

7.1 HIGH

AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:L

vuln.today AI

7.6 HIGH

PR:L confirmed by vendor CVSS; I:L added over official score because deployment resource modification constitutes an integrity impact not captured by I:N.

3.1 AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:L/A:L

4.0 AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:L/VA:L/SC:N/SI:N/SA:N

Primary rating from Vendor (elastic).

CVSS VectorVendor: elastic

Attack Vector

Network

Attack Complexity

Low

Privileges Required

Low

User Interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

None

Availability

Low

Lifecycle Timeline

Analysis Generated

Aug 13, 2026 - 20:00 vuln.today

CVE Published

Aug 13, 2026 - 19:13 cve.org

HIGH 7.1

DescriptionCVE.org

Authorization Bypass Through User-Controlled Key (CWE-639) in Kibana can lead to unauthorized cross-space access via Accessing Functionality Not Properly Constrained by ACLs (CAPEC-1). The result is disclosure of inference output from a trained model in a different space that the user is not authorized to list, read, or use, which exposes the behavior of a model. The same pattern also reached the deployment stop and deployment update operations, allowing an active trained model deployment in another space to be stopped or to have its allocated resources altered.

AnalysisAI

Cross-space authorization bypass in Kibana allows authenticated users to access trained machine learning model inference outputs, and stop or modify active model deployments, in Kibana Spaces they are not authorized to access. The vulnerability arises because API endpoints for ML model inference and deployment management accept user-controlled resource identifiers without enforcing space-level ACL boundaries, enabling privilege escalation across the Space isolation boundary. No public exploit code has been identified and Elastic has released fixed versions 8.19.20, 9.4.5, and 9.5.1 per advisory ESA-2026-126.

Technical ContextAI

Kibana's Spaces feature provides logical tenant isolation, allowing organizations to partition dashboards, ML models, and other resources so that users in one Space cannot access resources in another. The vulnerable code paths govern ML trained model inference (read operations) and deployment management (stop and resource-allocation update operations). CWE-639 (Authorization Bypass Through User-Controlled Key) identifies the root cause: the application uses a caller-supplied model or deployment identifier as a lookup key but fails to validate that the resolved resource resides within a Space the caller is authorized to access, matching CAPEC-1 (Accessing Functionality Not Properly Constrained by ACLs). The affected CPE is cpe:2.3:a:elastic:kibana:*:*:*:*:*:*:*:*, covering all versions prior to the patched releases. The CVSS vector AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:N/A:L reflects a network-reachable API requiring only low-privilege Kibana authentication.

RemediationAI

The primary fix is to upgrade Kibana to version 8.19.20, 9.4.5, or 9.5.1 as documented in Elastic security advisory ESA-2026-126 (https://discuss.elastic.co/t/kibana-8-19-20-9-4-5-9-5-1-security-update-esa-2026-126/389530). For organizations unable to patch immediately, a compensating control is to restrict access to Kibana's trained model inference and deployment management API endpoints (typically under /api/ml/ and /_ml/ paths) at the reverse-proxy or network layer to only users and services that legitimately require ML model access. The trade-off is that this may disrupt legitimate ML workflows if access lists are not carefully scoped. Additionally, auditing Kibana Space membership to ensure only trusted, vetted users hold any authenticated access reduces the attacker pool. Note that disabling the Elastic ML feature entirely (if not in use) eliminates exposure to this specific vulnerability path without impacting core observability functionality.