Haripriya Harikumar

Dr. Haripriya Harikumar is a Research Fellow at the University of Manchester's Centre for AI Fundamentals, specializing in Adversarial Machine Learning, AI Security, and Privacy-Preserving Technologies. Previously, she held a Postdoctoral Research Fellowship at Deakin University's Applied AI Intelligence Institute. She earned her PhD from Deakin University's Pattern Recognition and Data Analytics group.

Her research focuses on enhancing the trustworthiness of machine learning models through adversarial defense mechanisms, differential privacy, and out-of-distribution detection. Key areas include mitigating backdoor attacks, improving dataset bias analysis, and securing neural networks against trojan attacks.

Her work contributes to UN Sustainable Development Goal 4 (Quality Education) through advancements in AI ethics and security. She has published widely in top-tier conferences like CVPR, NeurIPS, and ICML, with a focus on robust AI systems.

As a Program Committee member for major ML conferences and Principal Investigator in the MCAIF project (2021–2026), she leads efforts in trustworthy AI fundamentals. Her multidisciplinary approach bridges cybersecurity, healthcare analytics, and computational biology.