Goal Reached Thanks to every supporter — we hit 100%!

Goal: 1000 CNY · Raised: 1359 CNY

100%
Buy Us a Coffee

CVE-2026-18950— Odh-dashboard: odh-dashboard: confused-deputy privilege escalation via unchecked roleref in rolebinding creation

Quick assessment

Affected
Red Hat Red Hat OpenShift AI 2.25
Exploitation
No confirmed in-the-wild exploitation; assess based on exposure
Recommended action
Check the vendor advisory and references for a fixed version. If immediate upgrade is impossible, restrict exposure and increase monitoring.

Red Hat OpenShift AI是美国Red Hat公司的一款混合云人工智能平台，为数据科学家和开发者提供跨混合云环境构建、开发、训练、部署、服务和监控机器学习（ML）模型及 AI 应用的全生命周期工具。 Red Hat OpenShift AI存在安全漏洞，该漏洞源于系统未正确验证RoleBindings创建时的roleRef字段，允许用户指定任意角色（包括高权限的cluster-admin），可能导致权限提升，攻击者可在其命名空间内获得未授权的提升访问权限并可能持久控制系统。

CVSS 8.8 · High EPSS 0.52% · P42

Possible ATT&CK Techniques 1 AI

T1078.004 · Cloud Accounts

Affected Version Matrix 10

VendorProduct Version RangeStatus
Red Hat Red Hat OpenShift AI (RHOAI) any unaffected
any unaffected
any unaffected
any unaffected
any unaffected
any unaffected
any unaffected
Red Hat Red Hat OpenShift AI 2.25 1785940823< * unaffected
Red Hat Red Hat OpenShift AI 3.3 1786109683< * unaffected
Red Hat Red Hat OpenShift AI 3.4 1786109665< * unaffected
Get alerts for future matching vulnerabilities Log in to subscribe

I. Basic Information for CVE-2026-18950

Vulnerability Information
Have questions about the vulnerability? See if Shenlong's analysis helps!
View Shenlong Deep Dive ↗

Although we use advanced large model technology, its output may still contain inaccurate or outdated information.Shenlong tries to ensure data accuracy, but please verify and judge based on the actual situation.

Vulnerability Title
Odh-dashboard: odh-dashboard: confused-deputy privilege escalation via unchecked roleref in rolebinding creation
Source: CVE Program / CVE List V5
Vulnerability Description
A flaw was found in odh-dashboard. An authenticated user of the dashboard can exploit a vulnerability related to how RoleBindings are created. The system does not properly validate the `roleRef` field, allowing a user to specify an arbitrary role, including highly privileged ones like `cluster-admin`. This can lead to privilege escalation, where an attacker gains unauthorized elevated access within their namespace and potentially persistent control over the system.
Source: CVE Program / CVE List V5
CVSS Information
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H
Source: CVE Program / CVE List V5
Vulnerability Type
特权授予不正确
Source: CVE Program / CVE List V5
Vulnerability Title
Red Hat OpenShift AI 安全漏洞
Source: CNNVD (China National Vulnerability Database)
Vulnerability Description
Red Hat OpenShift AI是美国Red Hat公司的一款混合云人工智能平台，为数据科学家和开发者提供跨混合云环境构建、开发、训练、部署、服务和监控机器学习（ML）模型及 AI 应用的全生命周期工具。 Red Hat OpenShift AI存在安全漏洞，该漏洞源于系统未正确验证RoleBindings创建时的roleRef字段，允许用户指定任意角色（包括高权限的cluster-admin），可能导致权限提升，攻击者可在其命名空间内获得未授权的提升访问权限并可能持久控制系统。
Source: CNNVD (China National Vulnerability Database)
CVSS Information
N/A
Source: CNNVD (China National Vulnerability Database)
Vulnerability Type
N/A
Source: CNNVD (China National Vulnerability Database)

Affected Products

Vendor Product Affected Versions CPE Subscribe
Red Hat Red Hat OpenShift AI 2.25 1785940823 ~ * cpe:/a:redhat:openshift_ai:2.25::el9
Red Hat Red Hat OpenShift AI 3.3 1786109683 ~ * cpe:/a:redhat:openshift_ai:3.3::el9
Red Hat Red Hat OpenShift AI 3.4 1786109665 ~ * cpe:/a:redhat:openshift_ai:3.4::el9
Red Hat Red Hat OpenShift AI (RHOAI) - cpe:/a:redhat:openshift_ai
Red Hat Red Hat OpenShift AI (RHOAI) - cpe:/a:redhat:openshift_ai
Red Hat Red Hat OpenShift AI (RHOAI) - cpe:/a:redhat:openshift_ai
Red Hat Red Hat OpenShift AI (RHOAI) - cpe:/a:redhat:openshift_ai
Red Hat Red Hat OpenShift AI (RHOAI) - cpe:/a:redhat:openshift_ai
Red Hat Red Hat OpenShift AI (RHOAI) - cpe:/a:redhat:openshift_ai
Red Hat Red Hat OpenShift AI (RHOAI) - cpe:/a:redhat:openshift_ai

II. Public POCs for CVE-2026-18950

# POC Description Source Link Shenlong Link
AI-Generated POC Premium

No public POC found.

Login to generate AI POC

III. Intelligence Information for CVE-2026-18950

请登录查看更多情报信息。

Vendor Advisories for CVE-2026-18950 (5)

Same Patch Batch · Red Hat · 2026-08-10 · 28 CVEs total

CVE-2026-18948 9.9 CRITICAL Feast: feast: unsafe dill deserialization of registry-stored udfs — rce on feature server
CVE-2026-14450 9.9 CRITICAL Maas-billing: maas api: privilege escalation via forged http headers due to missing authen
CVE-2026-18949 8.8 HIGH Odh-dashboard: odh-dashboard: clusterrole grants cluster-wide crud on secrets and rbac man
CVE-2026-18951 8.8 HIGH Odh-training-operator-rhel9: [trainer v2 security] trn-02: rhoai overlay aggregates trainj
CVE-2026-18617 8.8 HIGH Data-science-pipelines-operator: dspo: mysql dsn parameter injection via customextraparams
CVE-2026-18982 8.8 HIGH Odh-training-operator-rhel9: rhoai fork aggregates training job create onto native edit/ad
CVE-2026-18608 8.7 HIGH Data-science-pipelines-operator: dspo: operator clusterrole grants pods/exec:*, kubeflow.o
CVE-2026-18947 8.5 HIGH Feast: feast: authorization bypass in /materialize endpoints enables dos via unauthorized
CVE-2026-71576 8.5 HIGH Multicluster-global-hub: multicluster-global-hub: manager trusts self-asserted evt.source(
CVE-2026-15467 8.1 HIGH Trustyai-service-operator: trustyai-service-operator: lmevaljob sidecar containers bypass
CVE-2026-15581 8.0 HIGH Trustyai-service-operator: trustyai-service-operator: tas internal service bypasses kube-r
CVE-2026-63622 7.8 HIGH Libvirt: swtpm privilege escalation via symlink following
CVE-2026-18941 7.7 HIGH Feast: feast-operator: feast: default authentication mode is no_auth — shared multi-tenant
CVE-2026-18621 7.6 HIGH Data-sciences-pipeline: dsp: v1 argo template path accepts arbitrary workflow spec, bypass
CVE-2026-19387 7.6 HIGH Gstreamer: gstreamer1-plugins-bad-free: gstreamer: heap out-of-bounds write in adpcmdec im
CVE-2026-18611 7.5 HIGH Data-science-pipelines-operator: dspo: cryptographically weak secret generation (math/rand
CVE-2026-18618 7.5 HIGH Ml-metdata: bundled grpc 1.46.3 (2022) with published http/2 dos cves — directly reachable
CVE-2026-19389 7.1 HIGH Gstreamer: gstreamer1-plugins-ugly-free: gstreamer: integer overflow/underflow in asfdemux
CVE-2026-18620 7.1 HIGH Data-sciences-pipeline: user-controlled serviceaccount for workflow pods without authoriza
CVE-2026-19278 6.8 MEDIUM Stackrox: stackrox: privilege escalation via unanchored regular expressions in auth m2m ro

Showing top 20 of 28 CVEs. View all on vendor page &rarr; →

IV. Related Vulnerabilities

Same product: Red Hat OpenShift AI 2.25
Same vendor: Red Hat
Same weakness: CWE-266

V. Comments for CVE-2026-18950

No comments yet

Leave a comment