In December, I had the pleasure of participating in an expert panel addressing AI testing at the International Conference on Artificial Intelligence at Peterhouse College, at the University of Cambridge. You might be wondering what brought a cybersecurity researcher to an AI-centric event. I had the same scepticism when my multi-university research group decided we conduct AI-related research; what would my contribution be?
Our work has focused on defining meta-data for AI provenance, contributing to advancements in various facets of AI, including testing and auditability. Specifically, my focus lies on evaluating the dimensions of risk and trust within this context. Given the widespread impact of AI across diverse domains, there is a compelling opportunity for multidisciplinary research, consecutively, our group, has diverse expertise ranging from machine learning to psychology.
An initial publication on our work can be found here.
Author Dr Andrew M’manga