AI Security and Risk Management Training in Thimphu, Bhutan

This training provides professionals with foundational and practical knowledge of AI security, risk, and compliance. Participants will learn how to identify, assess, and manage security risks in AI systems. The course covers threats related to data, models, and deployment, along with governance frameworks and best practices to ensure safe and ethical use of AI.

Understanding AI Risks and Threats

Introduction to AI Security and Risk

• What makes AI systems vulnerable?
• Overview of AI lifecycle and threat surfaces
• Real-world AI failure and attack examples

Data Security in AI Systems

• Data poisoning and training data risks
• Ensuring data privacy and integrity
• Secure data pipelines and labeling practices

Model Risks and Vulnerabilities

• Adversarial attacks: examples and mechanisms
• Model inversion and membership inference attacks
• Defense techniques: robustness, noise filtering, adversarial training

Securing AI and Managing Risk

Secure AI Deployment and Monitoring

• Secure APIs and model access controls
• Model drift and continuous validation
• Logging, alerting, and rollback plans

AI Governance and Compliance

• Overview of AI regulations: EU AI Act, NIST AI RMF
• AI governance frameworks and documentation
• Model audits, bias detection, and explainability

Organizational Risk Management Practices

• Creating an AI risk management policy
• Roles and responsibilities for AI safety
• Risk registers and regular review cycles

Hands-On Exercise – Risk Assessment Walkthrough

• Analyze a sample AI use case (e.g., hiring, fraud detection)
• Identify risks, threats, and mitigations
• Build a mini risk register and review

Wrap-Up and Recommendations

• Summary of tools and frameworks
• Roadmap for secure and responsible AI adoption
• Final Q&A and feedback