CVE-2026-54640

OpenRemote before 1.26.2 contains an authentication bypass vulnerability in the console registration API that allows unauthenticated attackers to update existing console assets by supplying a known asset identifier.

In OpenRemote prior to 1.26.0, the datapoint export API constructs CSV crosstab SQL by concatenating asset and attribute names, allowing an authenticated user to inject SQL metacharacters. A crosstab export request can return injected SELECT output in the normal ZIP/CSV response, enabling database exfiltration. In multi-tenant deployments it may expose other tenants' data.

OpenRemote core (openremote-manager) before 1.31.0 does not perform realm or gateway-descendant authorization on inbound edge gateway events and maps event asset IDs only through the reversible GatewayConnector.mapAssetId transform. An attacker controlling a registered gateway can craft pre-image IDs in AttributeEvent or DELETE AssetEvent messages to overwrite attributes on, or delete, assets in any realm, breaking tenant isolation.

OpenRemote before 1.31.0 contains an authorization bypass vulnerability that allows realm admins to impersonate gateways in other realms because GatewayService only checks the gateway ID, not the realm. Attackers with WRITE:ADMIN in one realm can create a service user named after another realm's gateway ID to overwrite its descendant assets and open SSH tunnels into its local network.

OpenRemote 1.2.0 before 1.31.0 contains a missing authorization vulnerability in GatewayServiceResource that allows any authenticated user to start, stop and query gateway tunnels without an administrative role. Attackers can send POST or DELETE requests to the gateway tunnel endpoint to open persistent bidirectional TCP circuits into an edge gateway's network or tear down operator tunnels.

OpenRemote is an open-source internet-of-things platform. Prior to 1.24.2, the get, getUserClientRoles, and getUserRealmRoles methods authorize callers with the read:admin role without confirming that the target UUID belongs to the requested and authenticated realm. A realm administrator can retrieve user data from another tenant or the master realm, breaking tenant isolation.

OpenRemote prior to version 1.24.1 allows users with only read:assets privileges to write predicted datapoints via the predicted datapoint write endpoint. Version 1.24.1 fixes the issue.

OpenRemote before version 1.28.0 contains a cross-realm information disclosure vulnerability in the Notification REST API. Tenant administrators can read all tenants' sent notifications including message bodies by submitting a zero-parameter GET request to the notification endpoint.

OpenRemote notification deletion endpoints fail to enforce realm boundaries, allowing any realm administrator to delete notifications belonging to other realms. Attackers with write:admin role in one realm can send DELETE requests to remove notifications from the master realm or other tenants without authorization checks.

OpenRemote (org.openremote:openremote) versions up to 1.26.2 contain an insecure direct object reference (IDOR) vulnerability in the setAssetLinks endpoint of AlarmResourceImpl. The realm access check validates only a single realm obtained via realms.stream().findFirst() on a HashSet of realms from the request, rather than all realms. Due to non-deterministic HashSet iteration order, an authenticated attacker can, with roughly 50% probability per request (retryable), persist cross-tenant links and disclose victim asset names.