Track vulnerabilities like this one. Sign up free to get alerted when software you use is affected.
9.9

CVE-2026-112304: mindee/doctr can execute code via unsafe model loading

CVE-2026-112304 · published 1 month ago
Summary

The docTR library (mindee/doctr) up to version 1.1.0 loads pretrained model files without checking their safety. If an attacker provides a crafted model file, it can run arbitrary code on the system using the same rights as the program. Update to a newer release or only load model files from trusted sources to eliminate the risk.

What to do

The CVE record does not list a fixed version. Check the vendor's site or the advisory links below - a fix may already be released.

Affected software
VendorProductAffected versions
mindee doctr <= 1.1.0
Original advisory text
docTR through 1.1.0 Unsafe Deserialization RCE via load_pretrained_params()
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.
Fix within
Internet-facing 14 days
Internal At next upgrade
  • Not known to be exploited
  • Needs hands-on effort to exploit
  • Gives an attacker full control
Severity
9.9 Critical
Exploitation
<1% chance of attack within 30 days
Type
CWE-502Deserialization of Untrusted Data
Timeline
Published19 Oct 2026
Updated12 Nov 2026
First seen19 Oct 2026
Sources
CVE-2026-112304 · MITRE
Track software like this
Free during beta