Practical DevSecOps Logo
+1 (415) 800 4768

Certified AI Security Professional (CAISP)TM

The Certified AI Security Professional course offers an in-depth exploration of the risks associated with the AI supply chain, equipping you with the knowledge and skills to identify, assess, and mitigate these risks.

Through hands-on exercises in our labs, you will tackle various AI security challenges. You will work through scenarios involving model inversion, evasion attacks, and the risks of using publicly available datasets and models. The course also covers securing data pipelines, ensuring model integrity, and protecting AI infrastructure.

We start with an overview of the unique security risks in AI systems, including adversarial machine learning, data poisoning, and the misuse of AI technologies. Then, we delve into security concerns specific to different AI applications, such as natural language processing, computer vision, and autonomous systems.

In the final sections, you’ll map AI security risks against frameworks like the MITRE ATLAS (Adversarial Threat Landscape for Artificial-Intelligence Systems) and explore best practices for managing these risks. The course also covers secure AI development techniques, including differential privacy, federated learning, and robust AI model deployment.

By the end of this course, you will have a thorough understanding of the threats facing AI systems and strategies to secure them, ensuring the safe and ethical deployment of AI technologies in various industries.

After the training, you will be able to:

  • Earn the Certified AI Security Professional (CAISP) Certification by passing a 6-hour practical exam.
  • Prove to employers and peers, a practical understanding of the AI risks and mitigations.
Enroll Now
  • Self-paced Learning Mode
  • Browser-based Lab Access
  • 24/7 Instructor Support via Mattermost

Course Prerequisites

  1. Course participants should have knowledge of running basic Linux commands like ls, cd, mkdir, etc.,
  2. Familiarity with any scripting language like Python, Golang, or ruby helps. However, it’s not a necessity.

Learning Objectives

  1. Understand the critical role of AI security in protecting organizations from various threats.
  2. Identify the types of attacks targeting AI systems, including adversarial attacks, data poisoning, and model inversions.
  3. Develop strategies for assessing and mitigating security risks in AI models, data pipelines, and infrastructure.
  4. Apply best practices for securing AI systems, leveraging guidance from frameworks like MITRE ATLAS and other industry standards.

Chapter 1: Introduction to AI Security

  1. Course Introduction (About the course, syllabus, and how to approach it)
  2. About Certification and how to approach it
  3. Course Lab Environment
  4. Lifetime course support (Mattermost)
  5. An introduction to AI Security
  6. Basics of AI and ML
    1. What is AI?
    2. History and evolution of AI
    3. Key concepts in AI
      1. Intelligence
      2. Learning
      3. Adaptation
  7. Types of AI
    1. Narrow AI vs. General AI
    2. Supervised Learning: Learning from labeled data
    3. Unsupervised Learning: Finding patterns in unlabeled data
    4. Reinforcement Learning: Learning from feedback
    5. Natural Language Processing (NLP)
    6. Computer Vision
  8. Core Components of AI Systems
    1. Algorithms and Models: What they are and why they matter
    2. Data: Importance of quality data in AI
    3. Computing Power: Basic overview of hardware and software needs
  9. Introduction to Machine Learning
    1. What is Machine Learning?
    2. Differences between AI and ML
    3. Key ML concepts: Features (input variables) and Labels (output variables)
  10. Common Machine Learning Algorithms
    1. Simple explanations of Linear Regression and Decision Trees
    2. Introduction to Neural Networks
    3. Clustering basics (e.g., K-means)
  11. Applications of AI
    1. Natural Language Processing (NLP)
    2. Computer Vision
    3. Robotics
    4. Healthcare
  12. Importance of AI Security
    1. Potential Risks of AI
    2. Ethical Considerations in AI
    3. Regulatory and Compliance Requirements
  13. Threat Landscape in AI
    1. Types of AI Threats
    2. Case Studies of AI Security Incidents
  14. Basics of Deep Learning
    1. What is Deep Learning?
    2. Introduction to Neural Networks and how they work
    3. Brief overview of Convolutional Neural Networks (CNNs)
  15. Hands-on Exercise:
    1. Learn how to use our browser-based lab environment
    2. Setup Invoke AI a creative visual AI tool
    3. Create a chatbot with Python and Machine learning
    4. Text classification with TensorFlow
    5. Implementing Duckling for converting Text into Structured Data

Chapter 2: Attacking and Defending Large Language Models

  1. Introduction to Large Language Models
    1. Definition of Large Language Models
    2. How they work: Basics of language modeling
    3. Importance and impact of LLMs in AI
  2. Understanding LLM’s
    1. GPT (Generative Pre-trained Transformer)
    2. BERT (Bidirectional Encoder Representations from Transformers)
  3. Use Cases of LLMs
    1. Text Generation
    2. Text Understanding
    3. Conversational AI
    4. Language Translation
    5. Code Generation
  4. Security Concerns with LLMs
    1. Data Privacy
      1. Introduction to Data Privacy
      2. Risks to Data Privacy
      3. Mitigating Data Privacy Risks
    2. Model Theft
      1. Introduction to Model Theft
      2. Risks of Model Theft
      3. Preventing Model Theft
  5. Adversarial Attacks on LLMs
    1. Introduction to Adversarial Attacks
      1. What are adversarial attacks?
      2. How adversarial attacks exploit LLM vulnerabilities
    2. Types of Adversarial Attacks
      1. Evasion Attacks
      2. Poisoning attacks
    3. Defending Against Adversarial Attacks
      1. Techniques to enhance model robustness
      2. Best practices for training and deploying LLMs securely
  6. Real-World LLM attack tools on the internet
    1. XXXGPT
    2. WolfGPT
    3. WormGPT
    4. FraudGPT
  7. Hands-on Exercises:
    1. Scanning an LLM for agent based vulnerabilities
    2. Attacking AI Chat Bots
    3. Perform adversarial attacks using text attack
    4. Perform Webscraping using PyScrap
    5. Hide data in images using StegnoGAN
    6. Adversarial Robustness Toolbox (ART) – Python Library for Machine Learning Security – Evasion, Poisoning, Extraction, Inference – Red and Blue Teams
    7. Bias Auditing & “Correction” using aequitas

Chapter 3: LLM Top 10 Vulnerabilities

  1. Introduction to the OWASP Top 10 LLM attacks
    1. Purpose and Scope
    2. Purpose and Scope
  2. Model Inversion Attacks
    1. What are model inversion attacks?
    2. Reconstructing input data from model outputs for attacks
    3. Mitigation Strategies and best practices
  3. Adversarial Example Attacks
    1. What are adversarial examples?
    2. Deceiving models with small perturbations.
    3. Mitigation Strategies and best practices
  4. Data Poisoning
    1. What is data poisoning?
    2. Corrupting training datasets with malicious data
    3. Mitigation Strategies
      1. Techniques to detect and prevent data poisoning
      2. Ensuring data integrity in training processes
  5. Membership Inference Attacks
    1. Understanding membership inference
    2. Determining  if specific data was used in training
    3. Mitigation Strategies
      1. Approaches to protect against membership inference
      2. Ensuring model privacy and confidentiality
  6. Model Extraction Attacks
    1. Deep dive Model extraction attacks
    2. Attacking model functionality by replication
    3. Mitigation Strategies
  7. Model Misuse
    1. Factors to constitute model misuse?
    2. Consequences of using models for unintended purposes
    3. Mitigating model misuse
  8. Unintended Bias
    1. Defining unintended bias
    2. Manifestation of bias in AI models
    3. How to mitigate bias in AI models
  9. Lack of Transparency
    1. Transparency issues in AI
    2. Importance of model interpretability and explainability
    3. Improving transparency and best practices
  10. Unauthorized Model Access
    1. What is meant by unauthorized model access?
    2. Risks associated with unprotected model access
    3. Restricting unauthorized model access
  11. Ethical Concerns in LLM
    1. What are the ethical concerns in AI?
    2. Ethical dilemmas in LLM deployment
    3. Ethical Practices
  12. Hands-on Exercises:
    1. Prompt Injection
    2. Training Data Poisoning
    3. Excessive agency attack
    4. Adversarial attacks using foolbox
    5. Overreliance attack
    6. Insecure plugins
    7. Insecure output handling attack
    8. Exploiting Data Leakage
    9. Permission Issues in LLM

Chapter 4: AI Attacks on DevOps Teams

  1. Introduction to AI in DevOps
    1. Definition and principles of DevOps
    2. The role of AI in enhancing DevOps practices
  2. Types of AI attacks on DevOps
    1. Data Poisoning in CI/CD Pipelines
    2. Model Poisoning
    3. Adversarial Attacks
    4. Dependency Attacks
    5. Insider Attacks
  3. Real-World case of AI Attacks on DevOps
    1. Hugging Face artificial intelligence (AI) platform
    2. NotPetya attack
    3. SAP AI Core vulnerabilities
  4. Hands-on Exercises:
    1. Poisoned pipeline attack
    2. Dependency confusion attacks
    3. Exploitation of Automated Decision-Making Systems
    4. Compromising CI/CD infrastructure

Chapter 5: AI Threat Modelling

  1. Introduction to AI Threat Modelling
    1. Definition and purpose of threat modeling
    2. Importance in the context of AI security
  2. Key Concepts in AI Threat Modelling
    1. Assets
    2. Threats
    3. Vulnerabilities
    4. Attack Vectors
  3. AI Threat Modeling Methodologies
    1. STRIDE framework
    2. STRIDE GPT
    3. LINDDUN Framework
    4. MITRE ATLAS
  4. Tools for AI Threat Modelling
    1. Automated Threat Modelling Tools
    2. Manual Techniques
  5. Best Practices for AI Threat Modelling
  6. Hands-on Exercises:
    1. OWASP Threat Dragon
    2. Iruis Risk lab for threat modeling
    3. StrideGPT

Chapter 6: Supply Chain Attacks using AI

  1. An overview of the Supply Chain Security
  2. Introduction to AI Supply Chain Attacks
  3. Importance of Securing AI Supply Chains
  4. Software and Hardware Supply Chain Attacks
    1. Introduction of malicious code or hardware components
    2. Impact on the overall AI system integrity
  5. Mitigating AI Supply Chain Attacks
    1. Securing Data Pipelines
    2. Protecting Training Environments
    3. Managing Dependencies and Libraries
    4. Implementing Robust Access Controls
    5. Continuous Monitoring and Incident Response
    6. Adopting Secure Development Practices
    7. Collaborating with Supply Chain Partners
  6. Hands-on Exercises:
    1. Supply Chain Dependency Attack
    2. Transfer Learning and Pre-trained Model Attacks
    3. Flagging vulnerable dependencies using flag
    4. Model Inversion Attack
    5. Backdoor attacks using BackdoorBox

Chapter 7: Emerging Trends in AI Security

  1. Explainable AI (XAI)
    1. Importance of Explainability
    2. Techniques for Explainability
  2. AI Governance and Compliance
    1. Regulatory Requirements
    2. Best Practices for AI Governance
  3. Future Trends in AI Security
    1. Emerging Threats
    2. Innovations in AI Security
  4. Hands-on Exercises:
    1. Explainable AI basics
    2. AuditNLG to audit generative AI
    3. Scan malicious python packages using aura.

Practical DevSecOps Certification Process

  1. After completing the course, you can schedule the CAISP exam on your preferred date.
  2. Process of achieving Practical DevSecOps course certifications can be found on the exam and certification page.
Enroll Now
Back to Course List