What is the CVSS score of EUVD-2026-55826?

EUVD-2026-55826 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:L/A:N. EPSS exploitation probability: 0.3%.

Which versions of Data Science Pipelines are affected by EUVD-2026-55826?

Red Hat OpenShift AI's Data Science Pipelines component contains a privilege escalation vulnerability (CVE-2026-18620) that allows authenticated users to bypass authorization controls and gain access…. A patch is available.

How to fix or mitigate EUVD-2026-55826?

Within 24 hours, inventory all OpenShift AI deployments utilizing Data Science Pipelines and confirm multi-tenant configuration; document which external or internal users have authentication credentials. Within 7 days, enforce compensating controls: restrict all ServiceAccount RBAC roles to least-privilege bindings, implement network policies denying cross-namespace pod communication, enable detailed audit logging on CreateRun API calls with non-default ServiceAccount references, and apply Pod Security Policies to prevent privilege escalation. Within 30 days, establish a monitoring process for Red Hat OpenShift AI security advisories for CVE-2026-18620 patch release; evaluate upgrade feasibility to patched versions when available, and consider temporary operational measures such as dedicating isolated clusters for high-value workloads until patched.

Data Science Pipelines EUVD-2026-55826

| CVE-2026-18620 HIGH

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

2026-08-10 redhat

GHSA-g68c-wmxr-x3rh

Authentication Bypass Red Hat Openshift Ai Rhoai Red Hat

7.1

CVSS 3.1 · Vendor: redhat

Severity by source

Vendor (redhat) PRIMARY

7.1 HIGH

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

vuln.today AI

8.5 HIGH

Scope set to Changed because successful exploitation crosses tenant security boundaries into other users' pods and secrets; PR:L confirmed by required tenant credential.

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

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

Red Hat

7.1 MEDIUM

qualitative

Primary rating from Vendor (redhat).

CVSS VectorVendor: redhat

Attack Vector

Network

Attack Complexity

Low

Privileges Required

Low

User Interaction

None

Scope

Unchanged

Confidentiality

High

Integrity

Low

Availability

None

Lifecycle Timeline

Analysis Generated

Aug 10, 2026 - 21:29 vuln.today

CVE Published

Aug 10, 2026 - 20:45 nvd

HIGH 7.1

DescriptionCVE.org

A flaw was found in Data Science Pipelines. A restricted user, or tenant, can exploit an improper authorization vulnerability in the setDefaultServiceAccount function. By specifying a more privileged ServiceAccount (SA) during a CreateRun request, an attacker can bypass authorization checks. This allows the tenant to run their containers with elevated privileges, potentially leading to the disclosure of sensitive information (secrets) and the ability to execute commands within other users' pods.

AnalysisAI

Privilege escalation in Red Hat OpenShift AI's Data Science Pipelines component allows authenticated tenants to bypass authorization controls in the setDefaultServiceAccount function by submitting a crafted CreateRun request that references a more privileged ServiceAccount than they are entitled to use. Successful exploitation enables the tenant's pipeline containers to run with elevated Kubernetes RBAC permissions, resulting in cross-namespace secret disclosure and the ability to execute commands inside other tenants' pods. No public exploit code or active exploitation (CISA KEV) has been identified at time of analysis.

Technical ContextAI

The vulnerability resides in the Data Science Pipelines component of Red Hat OpenShift AI (RHOAI), identified by CPE cpe:2.3:a:red_hat:red_hat_openshift_ai_(rhoai). Data Science Pipelines is an MLOps orchestration layer built on top of Kubeflow Pipelines that schedules and executes containerized ML workloads on OpenShift/Kubernetes clusters, often in shared, multi-tenant configurations. The root cause is CWE-639 (Authorization Bypass Through User-Controlled Key): the setDefaultServiceAccount function accepts a user-supplied ServiceAccount identifier during CreateRun without validating that the requesting tenant holds authority to bind that SA. In Kubernetes, a pod's ServiceAccount directly determines its RBAC permissions; by attaching a privileged SA (e.g., one bound to cluster-admin or a cross-namespace role), the tenant's workload inherits those permissions, enabling it to enumerate and read secrets across namespaces and interact with the Kubernetes API as that SA.

RemediationAI

Consult the Red Hat advisory at https://access.redhat.com/security/cve/CVE-2026-18620 and apply any RHOAI update that addresses CVE-2026-18620; no exact patched version was confirmed in available data, so monitor the Bugzilla entry at https://bugzilla.redhat.com/show_bug.cgi?id=2510320 for fix version publication. As an interim compensating control, restrict CreateRun API access to the Data Science Pipelines endpoint to a minimal set of trusted users via OpenShift RBAC, reducing the population of accounts capable of triggering the flaw. Additionally, deploy an OPA/Gatekeeper or Kyverno admission policy that validates any ServiceAccount reference in pipeline run requests against a per-namespace allowlist of permitted SAs; this directly addresses the bypass mechanism without requiring a code patch, though it may require tuning to avoid breaking pipelines that dynamically select SAs. Audit existing pipeline run records for unauthorized SA references to detect any prior exploitation attempts. Treat any elevated-privilege SA binding discovered in tenant namespaces as a potential indicator of compromise.