Presenter
Maxwell Standen
Defence Science and Technology Group
Maxwell is a Senior Researcher at the Defence Science and Technology Group and a PhD Candidate in the School of Computer Science and Information Technology at Adelaide University. His research is at the intersection of Artificial Intelligence and Cyber Security, with a focus on developing robust and secure AI systems. His research has included adversarial machine learning, autonomous cyber operations, and multi-agent reinforcement learning.
Maxwell Standen from DSTG, will present a seminar on Thursday, 7 May 2026.
Title: Adversarial Machine Learning for Defence Applications: Securing Communication in Multi-Agent Systems.
Abstract: Multi-agent systems have the potential to address many complex real-world problems, particularly in a defence context such as counter UAV and cybersecurity. However, defence environments are often contested and so these systems must be secure and robust to adversarial interference. A significant attack surface for multi-agent systems are the communications between agents, which, while necessary for coordinating actions and sharing information, exposes the system to potential attacks.
This presentation discusses ongoing research in the field of adversarial machine learning, which aims to identify and mitigate the vulnerabilities that arise from inter-agent communication in multi-agent systems. This research includes techniques for finding the weakest links of the system, which improves attack effectiveness and allows a more accurate assessment of vulnerability. It also covers potential defences such as quantisation, which restricts the freedom of adversaries to interfere with messages. The research has the potential to significantly improve the security and robustness of multi-agent systems and is an important step towards operationalising such capability.
Click the button below to register, add an invitation to your calendar and join the seminar using the Teams/GovTeams link.
