Rembg: Unauthenticated SAM model injection permits local-file disclosure through malicious ONNX graphs

影響: pip:rembg < 2.0.82 → 2.0.82

The public /api/remove endpoint uses attacker-controlled extras.sam_model in both SAM filenames and GitHub download URLs. Path traversal plus a URL fragment lets an unauthenticated attacker place malicious ONNX models beside a target file. The models read the file as ONNX external data and encode its bytes into the returned PNG. This issue affects SAM's sam_model in 2.0.81 and returns complete file contents. Screenshot 1 — First unauthenticated request primes the path and returns HTTP 500. Tech…

GitHub Advisory Database에서 읽기

2026. 10. 27. 오후 11:45, GitHub Advisory Database에서 "Rembg: Unauthenticated SAM model injection permits local-file disclosure through malicious ONNX graphs"을(를) 게시했습니다 (유형: 취약점 권고). 심각도는 "높음"(high)입니다. 관련 기술: Python. CVE: CVE-2026-115924.