CVE-2026-34445
Due to this, an attacker could craft a malicious model that overwrites internal object properties.
Does this matter?
High impact if exploited, but EPSS currently rates exploitation as unlikely (0.29%). Schedule it in the normal patch cycle and watch for a rise in EPSS or a public exploit.
Description
Open Neural Network Exchange (ONNX) is an open standard for machine learning interoperability. Prior to version 1.21.0, the ExternalDataInfo class in ONNX was using Python’s setattr() function to load metadata (like file paths or data lengths) directly from an ONNX model file. It didn’t check if the "keys" in the file were valid. Due to this, an attacker could craft a malicious model that overwrites internal object properties. This issue has been patched in version 1.21.0.
- CVSS 3.1
- 8.6 HIGHCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H
- EPSS
- 0.29% probability · 21th percentile
- Public exploits
- None in Exploit-DB
- CISA KEV
- Not listed
- Weakness
- CWE-20, CWE-400, CWE-915
- Affected
- linuxfoundation/onnx
- Source
- security-advisories@github.com
EPSS 0.3% since 26 October 2026; no change recorded yet. Points are recorded when the score first appears and whenever it moves by a percentage point or more.
References
- https://github.com/onnx/onnx/commit/e30c6935d67cc3eca2fa284e37248e7c0036c46bPatch
- https://github.com/onnx/onnx/pull/7751Issue Tracking, Patch
- https://github.com/onnx/onnx/security/advisories/GHSA-538c-55jv-c5g9Patch, Vendor Advisory
Source: NVD record, EPSS from FIRST.org, KEV from CISA, exploits from Exploit-DB. Refreshed daily. Download this record as JSON.