CVE-2026-9800: Keycloak Policy Enforcer Authorization Bypass
A vulnerability in Keycloak's Policy Enforcer allows any logged-in user to bypass authorization controls by manipulating the request URL. An attacker can trick the system into accepting requests that should be blocked by including the configured access-denied page path as part of the URL—either as a path segment or query parameter. This circumvents role-based access control, scope restrictions, and User-Managed Access permissions, granting unauthorized access to protected resources. The flaw requires an attacker to already have some form of authentication, but the bypass is otherwise trivial to exploit once discovered.
Source data · NVD / CISA · public domain
- CVSS
- 3.1 · 8.1 HIGH · CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N
- Weaknesses (CWE)
- CWE-1025
- Affected products
- 1 configuration(s)
- Published / Modified
- 2026-06-25 / 2026-07-30
NVD description (verbatim)
A flaw was found in Keycloak Policy Enforcer. This vulnerability allows any authenticated user to bypass all authorization policies, including role, scope, and User-Managed Access (UMA) permission checks. By including the configured access-denied page path within a request URL, either as a path segment or a query parameter, an attacker can gain unauthorized access to protected resources.
13 reference(s) · View on NVD →
SEC.co analysis · AI-assisted, reviewed against source
Technical summary
The vulnerability exists in how Keycloak's Policy Enforcer handles authorization policy evaluation. The enforcer uses a configured access-denied page path as part of its policy logic, but fails to properly validate or sanitize this path when it appears in incoming requests. By including this path as a URL segment or query parameter, an authenticated attacker can cause the enforcer to skip or bypass policy checks—including role-based access control (RBAC), scope validation, and User-Managed Access (UMA) permission verification. The root cause is classified under CWE-1025 (Comparison Using Wrong Factors), suggesting the enforcer is comparing request paths incorrectly or incompletely. This allows the attacker to access resources that should be protected by these policies.
Business impact
Organizations deploying Keycloak as an identity and access management (IAM) solution face a critical breach of their authorization model. Any employee, contractor, or service account that has gained initial authentication can escalate their access to protected resources without proper authorization. This could expose sensitive data, allow unauthorized actions on critical systems, and bypass compliance controls that depend on role and scope enforcement. The blast radius spans all applications protected by the compromised Keycloak instance. Data breaches, unauthorized transactions, and regulatory violations become possible depending on what the protected resources do. For organizations in regulated industries (healthcare, finance, government), this may trigger mandatory breach notification and investigation obligations.
Affected systems
Red Hat build of Keycloak is directly affected. Organizations running Keycloak for identity and access management, particularly those using Policy Enforcer for fine-grained authorization, are at risk. Any application or service relying on Keycloak's policy enforcement for access control—including applications using role-based access, scope-based access, or User-Managed Access (UMA) models—inherits this vulnerability. Organizations should verify whether they are running Keycloak in a configuration where Policy Enforcer is active and responsible for authorization decisions.
Exploitability
Exploitability is high. The attack requires only valid authentication credentials, which many organizations assume their users possess legitimately. The exploit itself is straightforward: append or include the access-denied page path in a request URL. No special tooling or code execution is necessary; a simple HTTP request with the path manipulation is sufficient. No user interaction is required beyond the attacker making the malicious request. The main barrier is knowing the configured access-denied page path, which may be guessable or discoverable through reconnaissance. The CVSS vector (AV:N/AC:L/PR:L) reflects network accessibility and low attack complexity once authentication is obtained. The vulnerability is not currently listed on CISA's Known Exploited Vulnerabilities (KEV) catalog, but the straightforward nature of the attack means exploitation could emerge rapidly once public details circulate.
Remediation
Apply patches from Red Hat for the affected build of Keycloak. Verify the specific patch version against Red Hat's advisory, as the exact version depends on your current deployment branch. Organizations should prioritize patching Keycloak instances that enforce authorization policies for sensitive applications or data. In the interim, monitor access logs for suspicious requests containing the access-denied page path, implement additional network-level access controls if possible, and review Keycloak configurations to understand which applications depend on Policy Enforcer. Consider temporarily restricting authentication to trusted networks if the risk is judged critical.
Patch guidance
Contact Red Hat or consult their security advisories for the exact patch version applicable to your build of Keycloak. Red Hat typically releases patches for multiple active versions; identify your current version and apply the corresponding patch. Test patches in a non-production environment first to ensure no disruption to applications dependent on Keycloak policy enforcement. Patching should be treated as urgent given the ease of exploitation. Organizations should establish a change window for Keycloak patches that minimizes impact on dependent applications and plan for coordinated testing with application teams.
Detection guidance
Monitor Keycloak access and audit logs for requests that include the configured access-denied page path within the URL (as path segments or query parameters). Look for patterns where authenticated users are accessing resources outside their assigned roles or scopes—a successful exploitation attempt will show authorization bypass behavior. Review Policy Enforcer logs for skipped or incomplete policy evaluations. If you do not know your access-denied page path, check Keycloak's policy enforcer configuration in your deployment. Hunt for unusual authorization decisions that contradict the policies you have defined, especially if paired with suspicious request URLs. Correlate Keycloak logs with application logs to identify whether authorization decisions made by Keycloak align with actual access patterns in protected resources.
Why prioritize this
This vulnerability ranks HIGH (CVSS 8.1) and merits immediate attention because it directly undermines authorization—the core security boundary in any identity and access management system. The attack requires only authentication (a low bar in most organizations) and is trivial to execute. The impact is complete bypass of role, scope, and UMA permissions, affecting confidentiality and integrity of protected resources. It is not yet on the KEV catalog, but the simplicity of the exploit suggests public PoC code or active exploitation could emerge soon. Organizations relying on Keycloak for access control to sensitive data or systems should treat this as a critical remediation priority.
Risk score, explained
The CVSS score of 8.1 (HIGH) reflects: (1) network-accessible attack vector requiring no special position; (2) low attack complexity once authenticated; (3) requirement for valid user credentials (PR:L), a relatively low bar given the number of legitimate users most organizations have; (4) high impact to confidentiality (can access restricted data) and integrity (can perform unauthorized actions), and no impact to availability. The score does not account for the widespread deployment of Keycloak in identity-critical roles across many organizations, which further increases real-world risk. The vulnerability is a direct authorization bypass affecting the trust boundary between authentication and protected resources.
Frequently asked questions
Does an attacker need special privileges to exploit this?
No, they only need valid authentication—a regular employee or service account credential is sufficient. Once authenticated, the path manipulation can be done immediately without escalation. This makes the vulnerability particularly dangerous in organizations with many legitimate users.
Can this vulnerability be exploited externally, or only from inside the network?
The attack can be executed over the network (CVSS vector AV:N). However, the attacker must first obtain valid authentication credentials, which typically requires either being an authorized user, compromising user credentials, or exploiting other vulnerabilities. The path manipulation itself is then performed remotely.
What should we do if we cannot patch immediately?
Monitor Keycloak logs for requests containing your access-denied page path in suspicious contexts. Implement network-level controls to restrict access to Keycloak and dependent applications. Review your Policy Enforcer configuration and consider temporarily disabling non-critical policies if absolutely necessary. Establish an expedited patch schedule; this vulnerability should not remain unpatched for extended periods.
How can we tell if we have been exploited?
Search Keycloak audit and access logs for requests that include the configured access-denied page path in the URL, paired with authorization decisions that do not match your policies. Cross-reference with application logs to see if unauthorized users accessed resources they should not have. Look for accounts making requests outside their normal role or scope patterns. A forensic timeline may be necessary for full impact assessment.
This analysis is provided for informational purposes to assist security teams in risk assessment and remediation planning. Verify all technical details, affected versions, and patch availability against official Red Hat security advisories and vendor documentation. CVSS scores and vulnerability classifications are subject to interpretation and may be updated by vendors or CISA. This vulnerability is not currently listed on CISA's Known Exploited Vulnerabilities catalog as of the publication date, but public exploitation could emerge. Organizations are responsible for assessing their own environment, testing patches before deployment, and implementing appropriate controls. Consult with Red Hat support or your Keycloak vendor for guidance specific to your deployment configuration and version. Source: NVD (public-domain), retrieved 2026-08-03. Analysis generated by SEC.co (claude-haiku-4-5).
Related vulnerabilities
- CVE-2026-48860MEDIUMErlang/OTP SSL inet_tls_dist IP Spoofing Allows Cluster Bypass
- CVE-2023-54365HIGHTraefik HTTP/2 Denial of Service Vulnerability – Rapid Reset Attack
- CVE-2026-12112HIGHForeman MCP Server Session Hijacking Vulnerability
- CVE-2026-13201HIGHKubeVirt Symlink Following in virt-handler Privilege Escalation
- CVE-2026-1784HIGHHAProxy Configuration Injection in Red Hat OpenShift Routes
- CVE-2026-41731HIGHSpring for Apache Kafka Header Deserialization RCE (CVSS 8.1)
- CVE-2026-42965HIGHOpenShift Router SSRF via FQDN EndpointSlice
- CVE-2026-46579HIGHOpenShift Router Header Spoofing Bypasses Mutual TLS Authentication