AI/ML Penetration Testing

The AI/ML Penetration Test is based on the latest OWASP Machine Learning Security project, which covers the most common and high impact security issues known for machine learning systems. We further fuel the quality of our testing to identify complex issues using a manual approach.

AI/ML Penetration Testing Methodology

• Input Manipulation Attack
• Data Poisoning Attack
• Model Inversion Attack
• Membership Inference Attack
• Model Theft
• AI Supply Chain Attacks
• Transfer Learning Attack
• Model Skewing
• Output Integrity Attack
• Model Poisoning