Description

OpenRemote versions before 1.28.0 contain a cross-realm information disclosure vulnerability in the Notification REST API that allows per-realm tenant administrators to read all tenants' sent notifications including message bodies. Attackers with read:admin credentials in one realm can submit a zero-parameter GET request to the notification endpoint to retrieve sensitive notification metadata and message content from all realms.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "7.7",
  "pubDate": "2026-08-27T17:20:57.150Z",
  "pubdate": "2026-08-27T17:20:57.150Z",
  "executiveSummary": "OpenRemote versions prior to 1.28.0 are susceptible to a critical cross-realm information disclosure vulnerability residing within the Notification REST API.\nThis vulnerability stems from insufficient access control enforcement on multi-tenant architecture, allowing an authenticated per-realm tenant administrator to bypass logical isolation boundaries.\nAn attacker possessing valid 'read:admin' credentials within a single, isolated realm can query the notification endpoint to exfiltrate sensitive metadata and full message bodies belonging to other tenants hosted on the same instance.\nThe impact is significant, as it leads to a complete breach of inter-tenant confidentiality, exposing potentially sensitive user data, system notifications, and internal communications across the entire platform.\nThe exploitation process is low-complexity, requiring only an authenticated session with administrative read privileges and a simple, zero-parameter GET request, necessitating immediate mitigation to prevent unauthorized data access.",
  "technicalDetails": "The vulnerability is rooted in an authorization flaw within the Notification REST API of OpenRemote, where the backend fails to validate that a request initiated by a tenant administrator is scoped strictly to the resources belonging to their respective realm.\nIn a multi-tenant environment, the security model relies on the consistent application of tenant-based filtering during all database interactions. In this specific implementation, the API endpoint responsible for retrieving notifications performs a broad data fetch without applying the necessary WHERE clauses or context-based filtering required to distinguish between different organizational tenants.\nThe attack vector involves an authenticated actor with 'read:admin' level access to a specific realm. By submitting a crafted GET request to the notification endpoint—which requires no additional parameters or complex input—the actor triggers an unconstrained database query.\nBecause the underlying service layer does not enforce identity-based access control (ABAC) or role-based access control (RBAC) restrictions tied to the realm identifier, the backend processes the request as a global fetch operation.\nThe application subsequently retrieves the aggregate dataset of all notifications stored within the global system, including sensitive message bodies, timestamps, recipient information, and metadata pertaining to every configured tenant.\nThe server-side response encapsulates this cross-tenant data in a JSON payload, effectively bypassing the security perimeter intended to isolate sensitive tenant-specific information.\nThis flaw represents a failure in the application's authorization logic, where the system fails to verify the relationship between the authenticated user's session token and the resource-level ownership context. The network exposure is limited to authenticated users; however, the impact is catastrophic in shared-hosting environments where trust is partitioned by realm-level administrative roles. There are no specialized payloads required, as the vulnerability is inherent to the logic of the API request handling process itself.\nPost-exploitation, an attacker can enumerate, archive, and exfiltrate all system-wide notifications, facilitating large-scale data breaches and lateral reconnaissance within the platform's user base."
}