- Level: APPING-2
- Semester: S7
- Duration: 23h
- Language: French
- Teacher: Pierre Parrend
Summary
The development of artificial intelligence tools is profoundly transforming cybersecurity professions in every domain: monitoring, penetration testing, vulnerability and threat research, code and script generation, summarizing complex cases, and security response preparation.
This course first aims to define the main AI approaches — machine learning, language models — their use cases, and implementation best practices. It then provides useful tools for security operations, monitoring, and code and script generation.
Objectives
After completing this course, students will be able to:
- Apply machine learning algorithms for security
- Implement performance evaluation (detection and runtime) of learning algorithms
- Implement best practices for the use of language models
- Use language models for script and code generation
- Use language models for generating security procedures
- Evaluate the contribution, relevance, and limitations of using language models
Lecture outline
- Impact of AI on cybersecurity: defense and threats
- Impact of AI on the cognitive processes of security operators
- Use cases of artificial intelligence for cybersecurity
- Principles of machine learning
- Machine learning operations: classification, clustering, anomaly detection
- Application of machine learning for security
- Fundamentals of machine learning
- Principles of differentiable programming, generative AI, and language models
- Best practices for the use of language models
- Language models for security:
- Text generation (procedures and communication)
- Code generation (development, penetration testing)
- Code analysis and vulnerability research
- Ethical challenges of artificial intelligence
- Securing artificial intelligence
References
- Machine Learning and Security: Protecting Systems with Data and Algorithms, Clarence Chio, David Freeman, O’Reilly Media, 2018