Handbook

This course covers advanced AI technology for cyber security, including explainable AI and adversarial machine learning algorithms. Trustworthy AI for cyber security involves the development of artificial intelligence systems that can operate securely and reliably in a variety of contexts. These systems are designed to detect and respond to various threats in cyberspace, such as malware, phishing attacks, and advanced persistent threats. To accomplish this, trustworthy AI incorporates a range of security measures, including encryption, secure communication channels, intrusion detection systems, and continuous monitoring and updating of threat intelligence, explanation and robustness of models. In addition, trustworthiness is also ensured through robust testing and validation processes that ensure the reliability and consistency of AI systems, and the identification and mitigation of potential biases or unintended consequences.

In this course, students learn the mitigation strategies against adversarial machine learning and consider the ramifications of deep fakes. Students also learn the implications of explainable AI and how to achieve responsible AI models for offensive and defensive problems.  Students will also learn the threats and mitigation for large language models (including GPT models), and their impact on cyber threat hunting and threat intelligence, and penetration testing.