Code execution via untrusted attention-kernel configuration
Published May 24, 2026 · Updated May 24, 2026
Remote code execution in Hugging Face Transformers before 5.3.0 allows attackers to run Python code through a malicious model load. Configuration deserialization copies the untrusted _attn_implementation_internal value into a private attribute, and the hub-kernel path treats it as a repository identifier and loads unsandboxed code. Exploitation requires a user or application to load the attacker-controlled model with the standard from_pretrained API; the downloaded code runs with the loading process's OS privileges.
Summary
What happened
Remote code execution in Hugging Face Transformers before 5.3.0 allows attackers to run Python code through a malicious model load. Configuration deserialization copies the untrusted _attn_implementation_internal value into a private attribute, and the hub-kernel path treats it as a repository identifier and loads unsandboxed code. Exploitation requires a user or application to load the attacker-controlled model with the standard from_pretrained API; the downloaded code runs with the loading process's OS privileges.
The record
- CVE
- CVE-2026-4372
- Published
- May 24, 2026
- Updated
- May 24, 2026
- Vendor
- Hugging Face
- Product
- Transformers
- Classifications
- CWE-1066, CWE-502, T1204.002
- Attack vector
- local
- Privileges
- unauthenticated
Timeline
How it unfolded
- May 24, 2026CVE publishedPublication date reported by the CVE source.
- May 24, 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=unspecified <5.3.0
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
No public exploit references are available in this record.
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