The PhD research focuses on the fairness, explainability, and robustness of machine learning systems within the framework of causal counterfactual analysis using formalisms from probabilistic graphical models, probabilistic circuits, and structural causal models.