CyberDSA 2026 | Cyberstage Forum: Protecting Systems from Rogue AI Models and Agents

This session explains how to protect systems from rogue AI models and agents that may act outside their intended purpose. It focuses on controlling access, monitoring AI activities, reviewing outputs, requiring human approval for sensitive actions, and having emergency measures to stop or isolate unsafe AI behaviour. Keynote Address by Ir. Dr. Megat Zuhairy bin Megat Tajuddin Chief Executive of the National Cyber Security Agency Special Address by H.E. Dr. Mohamed Al Kuwaiti, Head of Cybersecurity for UAE Government, UAE Cybersecurity Council Speakers: 1. Mustafa Yasin Sheikh, Director General, National Communications Authority (NCA) of Somalia 2. Nazri Ahmad Zamani, Head of AI Unit, Cyber Risk Intelligence Unit, CyberSecurity Malaysia 3. Tajul Muhammad Taha bin Mohd Hanipiah, Chief Information Security Officer, MyDigital ID Moderator: Azrul Nazim Bin Abdul Aziz, Head of Industry Engagement and Collaboration, CyberSecurity Malaysia