Yolov13
Monthly
Code injection in the AutoBackend model-loading path of iMoonLab yolov13 (all versions up to commit 73289949533efac82bb5f72ec19b746618656bd2) allows a local, low-privileged attacker to execute arbitrary code by supplying or tampering with a model artifact whose imgsz, names, or kpt_shape metadata is passed to eval() in ultralytics/nn/autobackend.py. Exploitation is not remote: the attacker must already be able to plant or modify a model file that the victim's pipeline later loads (for example an untrusted downloaded checkpoint, a shared model registry, or a writable model directory), and the CVSS 4.0 vector confirms local access with low privileges (AV:L/PR:L). No public exploit code and no confirmed active exploitation were identified at time of analysis, and the upstream project was notified via GitHub issue #81 but has not responded, so no vendor-released patch exists as of this analysis.
Code injection in the AutoBackend model-loading path of iMoonLab yolov13 (all versions up to commit 73289949533efac82bb5f72ec19b746618656bd2) allows a local, low-privileged attacker to execute arbitrary code by supplying or tampering with a model artifact whose imgsz, names, or kpt_shape metadata is passed to eval() in ultralytics/nn/autobackend.py. Exploitation is not remote: the attacker must already be able to plant or modify a model file that the victim's pipeline later loads (for example an untrusted downloaded checkpoint, a shared model registry, or a writable model directory), and the CVSS 4.0 vector confirms local access with low privileges (AV:L/PR:L). No public exploit code and no confirmed active exploitation were identified at time of analysis, and the upstream project was notified via GitHub issue #81 but has not responded, so no vendor-released patch exists as of this analysis.