JeecgBoot CVE-2026-108607

Exploitation requires a valid authenticated JeecgBoot account (CVSS PR:L), after which the DELETE /airag/video/deleteVideoRecord endpoint accepts an attacker-supplied arbitrary userId without verifying ownership, and the companion GET /airag/video/listByUser endpoint returns other users' record ids without access checks. The attack targets the AI video generation feature of the jeecg-boot-module-airag module and affects Redis-stored video history, so deployments must have this AiRAG video functionality in use. Limiting factors: the attacker must already hold valid credentials (no unauthenticated access), impact is restricted to deleting another user's video history records one record per request (no confidentiality or system availability impact), and no special non-default configuration is otherwise required. Versions through 3.9.5 are affected.

This is a genuine but low-to-moderate priority issue, not a critical emergency. The CVSS 4.0 base score of 5.3 (Medium) is consistent with the vector CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:L/VA:N - network-reachable, low attack complexity, but requiring low-privilege authentication (PR:L) and yielding only limited integrity impact (VI:L) with no confidentiality or availability loss. The real-world risk is bounded by the impact: an attacker can delete other users' AI video generation history records one at a time, which is a cross-tenant data-integrity/tampering problem (CWE-639 IDOR), not code execution, data theft, or service outage. Signals are mostly aligned rather than conflicting: public exploit code exists (POC by VulnCheck), which raises the practical likelihood of opportunistic abuse, but there is NO CISA KEV listing and NO EPSS score provided in the input, so there is no evidence of active or widespread exploitation. The combination of 'POC available + authentication required + low impact' points to a credible but limited threat suitable for routine patch prioritization rather than emergency remediation. Note one metadata inconsistency to verify: the input tags this as 'Authentication Bypass,' but the mechanism (CWE-639, PR:L) is an authorization/access-control failure requiring a valid authenticated session - it is not an authentication bypass.

An authenticated low-privilege user of a JeecgBoot deployment calls the unchecked GET /airag/video/listByUser endpoint to enumerate other users' AI video record IDs, then sends DELETE /airag/video/deleteVideoRecord requests supplying a victim's arbitrary userId and record id to erase their Redis-stored video history, one record per request. A public Python proof-of-concept (published by VulnCheck via the github-poc reference) automates this, lowering the skill barrier; however, exploitation still requires a valid authenticated account and network access to the application.