How to meet it
OWASP AISVS v1.01
Supply Chain Security for Models · Model Artifact Integrity
Scan models for malicious code
Verify that models are scanned for malicious code before import.
OWASP AISVS v1.01, requirement 6.1.1
In plain English
Scan downloaded models for malicious code before they enter your environment. Model files can carry code that runs when they are loaded.
What good looks like
- Every imported model is scanned before it reaches internal storage.
- Files that fail are quarantined.
- Scan results are recorded with the model.
How teams usually do it
- Model scanning tools that inspect pickle and other formats for unsafe operations.
- Converting models to safe formats as part of import.
Evidence an auditor would accept
- Import pipeline configuration and scan reports.
Common mistakes
- Scanning only when a model is first downloaded, then pulling updates directly.
Not sure which requirements apply to your system? Take the assessment.
Requirement text from the OWASP AI Security Verification Standard, licensed under CC BY-SA 4.0. This guidance is shared under the same licence.
