Research Philosophy: Publishing Influential Research in AI‑Security

The Data Science Discipline of the School of EECS is hosting the following guest seminar:

Title: Research Philosophy: Publishing Influential Research in AI‑Security

Speaker: Prof Willy Susilo, University of Wollongong

Abstract: Artificial intelligence is rapidly reshaping how societies function, economies operate, and decisions are made at scale. As AI systems are increasingly integrated into critical infrastructure, autonomous platforms, healthcare, finance, and governance, securing these systems is no longer a purely technical challenge. It has become a foundational requirement for trust, safety, and long‑term societal resilience.

In this talk, I will share my research philosophy on how to develop and publish high‑quality, influential research in AI‑Security. The focus will be on research that not only meets the standards of leading international venues, but also shapes real‑world practice, informs policy, and influences large‑scale deployment. Drawing on selected examples from my work at the intersection of artificial intelligence and security, I will reflect on how impactful research emerges from identifying the right problems, asking the right questions, and sustaining a clear long‑term vision in a rapidly evolving field.

AI‑Security has become central to ensuring the robustness, trustworthiness, and accountability of intelligent systems. As models grow in scale, complexity, and autonomy, many traditional security assumptions no longer hold. Emerging threats such as adversarial manipulation, data poisoning, model extraction, misuse of generative systems, and systemic vulnerabilities challenge us to rethink security across the entire AI lifecycle, from data and model design through deployment and governance. Addressing these challenges requires not only deep technical innovation, but also strong strategic insight and interdisciplinary awareness.

Beyond technical contributions, this presentation will address a question of broad relevance to the research community. How can researchers navigate the fast‑moving AI‑Security landscape in order to produce work that truly matters? I will discuss guiding principles for building a coherent and enduring research agenda, recognising emerging inflection points, and ensuring that scholarly contributions translate into tangible improvements in the security and trustworthiness of AI systems deployed in the real world.