/

CVE-2026-81679

Severity:

HIGH RISK

(7.7)

Cross-Realm Info Disclosure (CWE-200) in OpenRemote

CVE-2026-81679

Deferred

Vulnerability type

Category:

Information Disclosure

CWE:

CWE-200

Impact:

Cross-Realm Data Exposure

Severity:

HIGH RISK

(7.7)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

OpenRemote's Notification REST API may disclose sent notifications across realm boundaries to a tenant administrator operating in only one realm. An attacker with read:admin credentials can request notification data and may receive sensitive metadata and message bodies belonging to other tenants.

Affected context

The vulnerability affects OpenRemote deployments using a Notification REST API configuration in which authorization is not sufficiently scoped to the requesting administrator's realm. It becomes risky when a tenant administrator has read:admin access and the notification endpoint retrieves records without enforcing the user's realm or tenant boundary.

Exploit preconditions

  • Attack surface: Notification REST API
  • Required access: Valid credentials with read:admin privileges in at least one realm
  • Trigger point: A GET request to the notification endpoint without restrictive query parameters
  • Why exploitation is possible: The API may return notification records across realm and tenant boundaries instead of applying authorization filters based on the authenticated administrator's scope

Why this happens

The underlying issue is insufficient access-control enforcement in a multi-tenant notification retrieval path. The API appears to authenticate the requester but does not consistently constrain returned records to the requester's authorized realm or tenant. This behavior maps to CWE-200: Exposure of Sensitive Information to an Unauthorized Actor, because authenticated users can obtain information outside their permitted data scope.

Potential impact

  • Attackers may read notification message bodies from other realms.
  • Exposed metadata could reveal tenant activity, users, events, or operational details.
  • Sensitive information contained in notifications could be disclosed to unauthorized administrators.
  • Organizations could face privacy violations, compliance exposure, or loss of tenant isolation.

General mitigation guidance

Mitigation generally focuses on enforcing authorization at the data-access layer rather than relying only on endpoint authentication.

Common defensive patterns include:

  • Applying mandatory realm and tenant filters to every notification query.
  • Denying broad collection retrieval unless the caller has an explicitly authorized administrative scope.
  • Enforcing object-level authorization for individual notification records.
  • Using secure-by-default API behavior that returns only the caller's permitted data.
  • Reducing trust in client-supplied realm, tenant, or filtering parameters by deriving scope from authenticated server-side identity.
  • Running authorization and cross-tenant isolation tests as part of API quality and security controls.

How teams detect issues like this

Organizations can look for this weakness through:

  • Static and semantic analysis of notification queries for missing realm or tenant predicates.
  • Code review checks that compare authenticated scopes with data-access filters.
  • API tests using administrators from separate realms to verify strict tenant isolation.
  • Monitoring for broad notification requests, unusually large response sets, and access to records outside the caller's normal realm.
  • Policy enforcement requiring object-level authorization for multi-tenant resources.
  • Automated security review tools that identify collection endpoints lacking explicit authorization constraints.

Sources

SourceLink
github.comgithub.com/openremote/openremote/security/advisories/GHSA-6ff4-4frc-r287
vulncheck.comvulncheck.com/advisories/openremote-before-1.28.0-cross-realm-information-disclosure-via-notification-api

Frequently asked

questions

Yes, where the Notification REST API is reachable and an attacker possesses valid credentials with the required administrative read scope. Exploitation does not necessarily require access to the underlying server or database.

The attacker needs read:admin credentials in at least one realm. The issue concerns insufficient realm and tenant authorization after authentication, rather than unauthenticated access.

Enforce server-side, object-level authorization and require every notification query to be scoped to the authenticated user's permitted realm and tenant boundaries.

Use deny-by-default authorization, derive tenant scope from trusted identity context, test cross-tenant access with separate accounts, and monitor for anomalous broad data retrieval.

Find out what's already
exploitable in your codebase.