High severity · published Apr 1, 2026
CVE-2026-34445
ONNX: Malicious ONNX models can crash servers by exploiting unprotected object settings.
8.6
CVSS
0.3%
EPSS
No
On KEV
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 vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:L/I:L/A:H
Attack vector
Network
Complexity
Low
Privileges
None
User interaction
None
Confidentiality
Low
Integrity
Low
Availability
High
Scope
Unchanged
Exploitation
Is it being used?
Exploit prediction (EPSS)0.3% · higher than 21% of CVEs
Exploited in the wildNo evidence yet
CISA KEVNot listed
Ransomware campaign useNot known
Affected products
Which versions
ON
onnx · onnx
< 1.21.0 · source: vendor, NVD
References
NVD
https://nvd.nist.gov/vuln/detail/CVE-2026-34445
CVE.org
https://www.cve.org/CVERecord?id=CVE-2026-34445
Patch
https://github.com/onnx/onnx/commit/e30c6935d67cc3eca2fa284e37248e7c0036c46b
Issue Tracking
https://github.com/onnx/onnx/pull/7751
Patch
https://github.com/onnx/onnx/security/advisories/GHSA-538c-55jv-c5g9
More from onnx
9.1
CVE-2026-28500
ONNX Untrusted Model Repository Warnings Suppressed by silent=True in onnx.hub.load() — Silent Supply-Chain Attack
9.1
CVE-2024-7776
A vulnerability in the `download_model` function of the onnx/onnx framework, before and including version 1.16.1, allows for arbitrary file overwrite due to inadequate prevention…
7.5
CVE-2026-27489
ONNX: Path Traversal via Symlink
7.1
CVE-2026-49114
ONNX symlink-following and path-traversal arbitrary file write
6.2
CVE-2026-115922
ONNX: Denial of service via unchecked expansion of packed sub-byte tensors in ONNX Python APIs