Practical AI Security: Attacks, Defenses, and Applications – eLearning

Build practical expertise in AI security and learn how modern AI- and LLM-based solutions can be developed, attacked, and secured.

Order e-learning course here

The course takes you from the fundamentals of machine learning to practical security techniques for generative AI and large language models. Through hands-on lab exercises, you will train models, build LLM-based applications and work through realistic red team scenarios.

You will explore both the offensive and defensive sides of AI security, including prompt injection, automated penetration testing agents, MCP attacks, AI threat modelling, secure code review, AI gateways and guardrails. The course also introduces tools and technologies such as Hugging Face, LangChain, OpenAI APIs, Claude, MCP and Google’s Secure AI Framework.

What will you gain from the course?

• Understand the difference between AI, machine learning and large language models
• Learn the fundamentals of neural networks, generative AI and the ML lifecycle
• Understand how large language models and transformer architecture work
• Develop practical prompt engineering skills
• Work with AI tools such as Hugging Face and LangChain
• Build AI applications and solutions using Retrieval-Augmented Generation
• Understand common attacks on AI and LLM-based systems
• Explore prompt injection and AI red teaming
• Work with Model Context Protocol servers and security tools
• Carry out AI threat modelling and secure code reviews
• Implement defensive measures such as AI gateways and guardrails

 Who is the course for?

The course is suitable for anyone interested in the use and security of AI in cybersecurity, especially if you want to understand both offensive and defensive methods for securing AI systems.

The course is at a beginner to intermediate level.

Prerequisites

You should have:

• Basic knowledge of cybersecurity and penetration testing
• An understanding of basic programming concepts
• Experience with a high-level language used in machine learning, such as Python or similar

Course content

• Fundamentals of AI, machine learning and neural networks
• Generative AI and large language models
• Transformer architecture, tokenisation and embeddings
• Prompt engineering and development of LLM-based applications
• Hugging Face, LangChain and APIs for AI development
• Prompt injection, jailbreaking and data exfiltration
• AI red teaming and automated security agents
• Model Context Protocol and security in MCP servers
• AI threat modelling with frameworks such as STRIDE, MITRE ATLAS and MAESTRO
• Secure code review, AI gateways and guardrails
• AI governance, compliance and securing AI supply chains
• Use of AI in Security Operations and reverse engineering