This paper introduces a framework for modeling fairness using causal inference, focusing on *counterfactual fairness*. Counterfactual fairness ensures that a decision is fair if it remains the same in both the actual world and a counterfactual world where the individual belongs to a different demographic group. The authors define this concept and demonstrate its application through a real-world problem of fair prediction of success in law school. They contrast counterfactual fairness with other fairness definitions and show that counterfactual fairness can be achieved through a causal model that captures the relationship between protected attributes and data. The paper also discusses the implications of counterfactual fairness, including how it addresses historical biases and incompatibility issues between fairness criteria. Finally, the authors present an algorithm for implementing counterfactual fairness and evaluate it on a practical dataset, showing that it trades off some accuracy for fairness.This paper introduces a framework for modeling fairness using causal inference, focusing on *counterfactual fairness*. Counterfactual fairness ensures that a decision is fair if it remains the same in both the actual world and a counterfactual world where the individual belongs to a different demographic group. The authors define this concept and demonstrate its application through a real-world problem of fair prediction of success in law school. They contrast counterfactual fairness with other fairness definitions and show that counterfactual fairness can be achieved through a causal model that captures the relationship between protected attributes and data. The paper also discusses the implications of counterfactual fairness, including how it addresses historical biases and incompatibility issues between fairness criteria. Finally, the authors present an algorithm for implementing counterfactual fairness and evaluate it on a practical dataset, showing that it trades off some accuracy for fairness.