Summary
Combodo iTop versions before 3.2.3 may apply silo-based access checks only to classes explicitly included in an OQL SELECT clause. Under conditions where a user can submit or influence such queries, this gap could expose records or related data outside the user's intended visibility scope.
Affected context
The issue occurs in iTop's OQL query processing when silo access controls are evaluated only for classes present in the SELECT clause. Deployments using silo restrictions and query paths that reference additional classes may be affected, particularly where users can construct, modify, or invoke OQL queries through exposed application functionality.
Exploit preconditions
- Attack surface: iTop web application functionality that processes OQL queries
- Required access: Access to an authenticated or otherwise permitted feature that accepts or invokes affected OQL queries
- Trigger point: Evaluation of an OQL statement involving classes not directly represented in the SELECT clause
- Why exploitation is possible: Silo authorization checks may not be applied consistently to all classes contributing data to the query result
Why this happens
The underlying bug is an incomplete authorization check in OQL query evaluation. Access control is applied based on the classes in the SELECT clause rather than comprehensively covering all classes involved in the query, allowing silo restrictions to be bypassed for certain referenced data. This behavior maps to CWE-200: Exposure of Sensitive Information to an Unauthorized Actor because information may be returned beyond the requesting user's authorized scope.
Potential impact
- Unauthorized users may view records restricted by silo policy
- Sensitive IT service, asset, or organizational data could be disclosed
- An attacker could correlate exposed records to infer additional business information
- The incident could increase privacy, compliance, and operational risks
General mitigation guidance
Mitigation generally centers on enforcing authorization consistently across every class and relationship participating in an OQL query.
Defensive patterns include:
- Applying silo and object-level access checks to all query-referenced classes, not only selected fields
- Using centralized authorization enforcement after query parsing and before result delivery
- Restricting user-controlled query construction and exposing narrowly scoped query interfaces where possible
- Returning only the minimum data required for the requested operation
- Running security tests for cross-silo access and relationship-based data exposure
- Applying least privilege to application roles and data access paths
How teams detect issues like this
Detection can include static and architectural review of OQL authorization paths, with attention to checks derived solely from SELECT targets. Code review should examine whether joins, subqueries, relationship traversals, and referenced classes receive equivalent silo enforcement. Dynamic testing can use separate test identities and attempt cross-silo queries, while policy enforcement can require authorization tests for every query execution path. Automated security review tools may help identify incomplete object-level authorization patterns.

