Select your report type
Model File Format
.keras, .pkl, .tensorflow, .joblib, .json...
This is for AI/ML model files, not source code. Attack vectors include deserialization flaws, enabling code execution, and neural backdoors that embed malicious behaviors within models.
up to $1,500 USD
Open Source Repository
huggingface / transformers, keras / keras...
This is for AI/ML apps and libraries. These vulnerabilities often found in OWASP categories include misconfigurations and injection attacks in GitHub codebases.
Submissions closed