CVE-2026-112304
CVE-2026-112304 · docTR through 1.1.0 Unsafe Deserialization RCE via load_pretrained_params()
This page contains facts from public vulnerability sources. Check the original advisory before updating.
docTR through 1.1.0 contains an unsafe deserialization vulnerability in load_pretrained_params() that loads checkpoints with torch.load without weights_only=True. Attackers who control a local or URL-supplied checkpoint can embed a malicious pickle payload to execute Python code with the loading process's privileges on torch releases before 2.6.
CVSS 7.7 · High · EPSS 0.3%
- Published
- Updated
Affected products and versions
doctr
- Versions
- ≥ 0, ≤ 1.1.0: Affected (custom)
No fix published
Original sources
- https://github.com/mindee/doctr/issues/2175
- https://github.com/mindee/doctr/commit/687503dd15592483c4699ee51280b41938579a9e
- https://github.com/mindee/doctr/blob/v1.1.0/doctr/models/utils/pytorch.py
- https://www.vulncheck.com/advisories/doctr-through-1.1.0-unsafe-deserialization-rce-via-load-pretrained-params
- https://nvd.nist.gov/vuln/detail/CVE-2026-112304