/

CVE-2026-33650

Severity:

HIGH RISK

(7.6)

Authorization Bypass / Incorrect Authorization

CVE-2026-33650

Analyzed

Vulnerability type

Category:

Authorization Bypass / Incorrect Authorization

CWE:

CWE-863

Impact:

Privilege Escalation and Unauthorized Content Management

Severity:

HIGH RISK

(7.6)

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

WWBN AVideo up to and including version 26.0 contains an authorization logic flaw where the "Videos Moderator" permission is incorrectly reused as a gate for full video editing and ownership changes.

A user granted this limited moderation role can escalate their privileges to transfer ownership and delete any video, exceeding the intended scope of their permissions.

Affected context

This vulnerability affects WWBN AVideo deployments where:

  • The "Videos Moderator" permission is assigned to non-administrator users
  • The videoAddNew.json.php endpoint is used for video editing and ownership changes
  • The videoDelete.json.php endpoint enforces ownership but not higher-level authorization

In this context, the moderation role is documented as limited to changing video publicity (Active, Inactive, Unlisted), but the backend reuses the same permission flag to authorize broader video management operations.

Exploit preconditions

  • Attack surface: Application web interface / HTTP API for video management
  • Required access: An authenticated user account with the "Videos Moderator" permission
  • Trigger points:
  • videoAddNew.json.php for editing videos and transferring ownership
  • videoDelete.json.php for deleting videos based on ownership
  • Why exploitation is possible:
  • Permissions::canModerateVideos() is used as an authorization gate for full video editing, including changing ownership, even though the role is documented as only controlling publicity.
  • videoDelete.json.php checks only whether the caller is the owner, not whether they have a higher-privileged role.
  • By first transferring ownership of a target video to themselves via the overly broad edit operation, a moderator can then pass the ownership check in the delete endpoint and remove any video.

Why this happens

The core issue is an incorrect authorization boundary where a single permission (canModerateVideos) is overloaded to guard operations that have different risk levels.

Instead of mapping each sensitive action (publicity change, metadata edit, ownership transfer, deletion) to a dedicated permission or clearly separated authorization path, the code treats all "edit-like" actions as equivalent.

This maps to CWE-863: Incorrect Authorization, because the system performs authorization checks that do not accurately reflect the intended security policy: a limited moderation role is allowed to execute high-privilege functionality.

The asymmetry between videoAddNew.json.php (broadly gated by canModerateVideos) and videoDelete.json.php (ownership-only check) enables a privilege escalation chain through ownership manipulation.

Potential impact

  • A user with only "Videos Moderator" permissions may:
  • Transfer ownership of any video to their own account
  • Delete videos owned by other users after taking ownership
  • Perform broader video management operations that should be reserved for administrators or owners
  • Unauthorized removal of legitimate content, potentially impacting:
  • Content availability and business continuity
  • Users' trust in the platform's access controls
  • Potential abuse in multi-tenant or role-separated environments where moderation roles are widely delegated

General mitigation guidance

Effective mitigation centers on separating permissions by operation and making authorization checks reflect the documented policy rather than convenience:

  • Introduce distinct authorization checks for high-risk actions such as:
  • Ownership transfer
  • Video deletion
  • Editing critical metadata

and avoid reusing a single "moderate" permission for all of them.

  • Align backend authorization with documented role capabilities so that:
  • A "Videos Moderator" role is strictly limited to non-destructive actions such as changing publicity status, if that is the stated intent.
  • Enforce defense in depth on sensitive endpoints:
  • Require both appropriate role-based permission and ownership (or a clearly defined delegation rule) for operations like deletion and ownership reassignment.
  • Normalize authorization logic:
  • Centralize permission checks in a well-reviewed policy layer instead of duplicating logic in individual scripts.
  • Apply least privilege to role definitions:
  • Ensure that no role has more effective power than its description implies, especially in environments where roles are delegated to non-admin staff.

How teams detect issues like this

Organizations can identify and monitor similar issues through a combination of review and automation:

  • Code review patterns:
  • Inspect endpoints that perform sensitive actions (delete, transfer, update ownership, change ACLs) and verify they reference an appropriate, specific permission.
  • Look for reuse of "broad" or generic permissions (e.g., canModerateVideos) across operations with very different risk profiles.
  • Check for asymmetry where different endpoints on the same resource use different and inconsistent authorization checks.
  • Static analysis and policy enforcement:
  • Define rules that flag calls to security-critical functions or endpoints that do not go through a centralized authorization layer.
  • Enforce architectural constraints that every controller method passes through a standard permission-checking interface.
  • Runtime and logging strategies:
  • Log ownership changes, deletions, and other high-impact operations with actor identity and role information.
  • Periodically review logs for patterns where non-admin roles perform a high volume of destructive or ownership-changing operations.
  • Automated review tools can be configured to highlight unauthorized usage of sensitive actions, especially where permissions appear misaligned with documented role capabilities.

Sources

SourceLink
github.comgithub.com/WWBN/AVideo/commit/838e16818c793779406ecbf34ebaeba9830e33f8
github.comgithub.com/WWBN/AVideo/security/advisories/GHSA-8x77-f38v-4m5j

Frequently asked

questions

Yes. Any remote user who can authenticate to the platform and is granted the "Videos Moderator" permission can exploit this issue through the web interface or corresponding APIs, without needing local system access.

The attacker must have a valid account with the "Videos Moderator" permission. Users without this role, or anonymous users, cannot directly exploit the described privilege escalation chain.

The safest pattern is to clearly separate permissions for different classes of operations, especially destructive or ownership-related ones, and ensure every endpoint checks the most restrictive appropriate permission. Centralizing authorization logic and mapping it explicitly to business requirements helps prevent misalignment between intended and actual capabilities.

Adopt a rigorous access-control design process:

  • Define roles and permissions from a business perspective first, then implement them one-to-one.
  • Use a centralized authorization layer so that all sensitive actions share consistent, testable policies.
  • Include authorization scenarios in threat models and security reviews, and add automated tests that verify that limited roles cannot perform admin-level operations.
Find out what's already
exploitable in your codebase.