Reviews: Algorithmic Assurance: An Active Approach to Algorithmic Testing using Bayesian Optimisation

This paper presents a method to find the highest divergence between an ML model their goals by mapping the problem to a Bayesian optimization problem. The idea is very original and I found the paper very creative. The text is easy to follow. Theorem I is not properly defined. The proof is based on an example, which the authors claim can be easily generalized, but they do not provide such generalization.