Local code execution via unsafe pt2 deserialization
Published Mar 22, 2026 · Updated Mar 22, 2026
Unsafe deserialization in PyTorch 2.10.0 allows local users to execute code by loading a crafted .pt2 model archive. The torch.export.load path trusts use_pickle in model_weights_config.json, and deserialize_torch_artifact can fall back to pickle with weights_only disabled. Reachability requires an application to pass an attacker-controlled archive to the loader, after which pickle reducers execute with the process's privileges.
Summary
What happened
Unsafe deserialization in PyTorch 2.10.0 allows local users to execute code by loading a crafted .pt2 model archive. The torch.export.load path trusts use_pickle in model_weights_config.json, and deserialize_torch_artifact can fall back to pickle with weights_only disabled. Reachability requires an application to pass an attacker-controlled archive to the loader, after which pickle reducers execute with the process's privileges.
The record
- CVE
- CVE-2026-4538
- Published
- Mar 22, 2026
- Updated
- Mar 22, 2026
- Vendor
- PyTorch
- Product
- PyTorch
- Classifications
- CWE-20, CWE-502, T1204.002
- Attack vector
- local
- Privileges
- authenticated
Timeline
How it unfolded
- Mar 22, 2026CVE publishedPublication date reported by the CVE source.
- Mar 22, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=2.10.0
What conditions does exploitation require?
What is affected?
Published CVSS scores
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- Crafted .pt2 archive proof of conceptproof of concept · demonstrated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo