Description

AI Security and Responsible AI Practices

Ethical development and responsible deployment of AI and ML systems.

Learn the latest technology in AI and ML security to safeguard against AI attackers and ensure data integrity and user privacy.
Navigate privacy and ethical considerations to gain insights into responsible AI practices and address ethical consideration.
Explore emerging trends and future directions in AI, ML, security, ethics, and privacy focusing on key concepts including threats, vulnerabilities, and attack vectors.
Recognize and understand the privacy aspects of AI and ML, including data protection, anonymization, and regulatory compliance
Get the essential skills to protect your AI system against cyber attacks. Explore how generative AI and LLMs can be harnessed to secure your projects and organizations against AI cyber threats. Develop secure and ethical systems while being mindful of privacy concerns with real-life examples that we use on a daily-basis with ChatGPT, GitHub Co-pilot, DALL-E, Midjourney, DreamStudio (Stable Diffusion), and others. Gain a solid foundation in AI and ML principles and be better prepared to develop secure and ethical systems while being mindful of privacy concerns. Authors Omar Santos and Dr. Petar Radanliev are industry experts to guide and boost your AI security knowledge.

Table of contents
Introduction
AI Security and Responsible AI Practices: Introduction
Module 1: Fundamentals of AI and ML
Module introduction
Lesson 1: Overview of AI and ML Implementations
Learning objectives
1.1 Delving into supervised, unsupervised, and reinforcement learning
1.2 Diving into applications and use cases
1.3 Strategies in preprocessing and feature engineering
1.4 Navigating through popular and traditional ML algorithms
1.5 Exploring model evaluation and validation
Lesson 2: Generative AI and Large Language Models (LLMs)
Learning objectives
2.1 Introduction to generative AI
2.2 Delving into large language models (LLMs)
2.3 Exploring examples of AI applications we use on a daily basis
2.4 Going beyond ChatGPT, MidJourney, LLaMA
2.5 Exploring Hugging Face, LangChain Hub, and other AI model and dataset sharing hubs
2.6 Modern AI model training environments
2.7 Introducing LangChain, templates, and agents
2.8 Fine tuning AI Models using LoRA and QLoRA
2.9 Introducing retrieval-augmented generation (RAG)
Module 2: AI and ML Security
Module introduction
Lesson 3: Fundamentals of AI and ML Security
Learning objectives
3.1 Importance of security in AI and ML systems
3.2 OWASP top 10 risks for LLM applications
3.3 Exploring prompt injection attacks
3.4 Surveying data poisoning attacks
3.5 Understanding insecure output handling
3.6 Discussing insecure plugin design
3.7 Understanding excessive agency
3.8 Exploring model theft attacks
3.9 Understanding overreliance of AI systems
Lesson 4: How Attackers Are Using AI to Perform Attacks
Learning objectives
4.1 Exploring the MITRE ATLAS framework
4.2 AI supply chain security
4.3 Automated vulnerability discovery and creating exploits at scale
4.4 Intelligent data harvesting, OSINT, automating phishing, and social engineering attacks
4.5 Exploring examples of deepfakes and synthetic media
4.6 Dynamic obfuscation of attack vectors
Lesson 5: AI System and Infrastructure Security
Learning objectives
5.1 Secure development practices
5.2 Monitoring and auditing
5.3 Software Bill of Materials (SBOMs) and AI Bill of Materials (AI BOMs)
5.4 Using CSAF and VEX to accelerate vulnerability management
Module 3: Privacy and Ethical Considerations
Module introduction
Lesson 6: Privacy and AI Fundamentals
Learning objectives
6.1 Understanding key privacy considerations in AI implementations
6.2 Bias and fairness in AI and ML systems
6.3 Transparency and accountability
6.4 Understanding differential privacy
6.5 Exploring secure multi-party computation (SMPC)
6.6 Understanding homomorphic encryption
6.7 Understanding the AI data lifecycle management
6.8 Delving into federated learning
Lesson 7: AI Ethics
Learning objectives
7.1 Ethical considerations in AI development
7.2 Responsible AI frameworks
7.3 Policy frameworks
7.4 Exploring strategies to mitigate bias
Lesson 8: Legal and Regulatory Compliance
Learning objectives
8.1 Overview of upcoming regulations and guidelines
8.2 Ensuring compliance in AI and ML systems
8.3 Case studies and best practices
Summary
AI Security and Responsible AI Practices: Summary