MINT Seminar Dec 5: Harriet Farlow and Tania Sadhani on AI Security Likelihood Analysis

Abstract: Understanding AI risk means looking at both severity and likelihood—and while we’ve started to grasp the potential severity of AI incidents, likelihood remains a crucial piece of the puzzle. Our recently open-sourced likelihood analysis framework is designed to fill this gap, helping estimate how often different types of AI incidents might occur. This framework bridges short-term risks like security vulnerabilities and misuse with long-term concerns around safety and alignment, providing a more complete picture of AI risk. Developed in collaboration with ANU MINT Lab and UNSW, and funded by Foresight, this project is led by Mileva Security Labs. By combining real-world data and Bayesian modelling, we aim to give organisations practical tools to prioritise mitigations. We’d love your feedback to refine this work and make it even more impactful!

Bio: Harriet Farlow is the founder of Mileva Security Labs, specialising in AI and cybersecurity. She is a PhD candidate at UNSW Canberra, with a focus on adversarial machine learning and AI risk quantification. Harriet’s work bridges technical research and policy, and her open-source projects aim to empower organisations to navigate the complex landscape of AI security. She has spoken at DEF CON and other leading forums, advocating for practical and scalable AI risk solutions.Tania Sadhani

Tania Sadhani is an AI security researcher at Mileva Security Labs and an Honours student in Machine Learning at ANU. With a strong focus on adversarial threats and AI misuse, Tania contributes to cutting-edge research on AI risk and security frameworks. She is passionate about advancing methodologies that ensure AI systems are both safe and resilient.