About This Training
This intensive 8-hour deep-dive transforms you from an AI observer into an AI adversary—and ultimately, an AI defender. You'll think, act, and attack like a real threat actor, learning to exploit the same vulnerabilities that keep security teams awake at night.
Hands-on Experience
You will gain hands-on experience in:
- Mapping AI system architectures and identify attack surfaces across the entire ML pipeline
- Crafting sophisticated prompt injections that bypass safety guardrails
- Planting undetectable backdoors in neural networks
- Leveraging RAG systems and AI agents to access sensitive data
Training Format
Lectures
20-40% of the training
Hands-on Labs
60-80% of the training
Who Should Take This Course
- Security engineers
- Software engineers
- Data scientists
- ML engineers
- Ops engineers
- Anyone who is:
- Willing to expand their knowledge beyond traditional pentesting into AI-specific threats
- Wanting to understand how their creations can be weaponized
Student Requirements
A laptop with a browser (Chrome or Firefox recommended)
We have online cloud labs that require a browser capable of web socket connections.
Note: Even iPads would be sufficient for this training.
Instructor
Marudhamaran Gunasekaran
Principal Security Consultant
Practical DevSecOps
Marudhamaran Gunasekaran is a Security Consultant with a strong passion for securing software development through training and consulting. He is the developer and maintainer of the OWASP ZAP Dot Net API and speaks at various meetups and conferences on Agile Software Development and Security.
His certifications include Azure Certified Security Engineer, Microsoft Certified Trainer, ISO 27001 Lead Auditor, Professional Scrum Master I/II/III, and Certified DevSecOps Professional.