The European Union's General Data Protection Regulation (GDPR) introduces new rules for algorithmic decision-making, including a "right to explanation." This regulation, set to take effect in 2018, restricts automated decisions that significantly affect individuals, particularly those based on user-level predictors. It also creates a right for users to request explanations for algorithmic decisions made about them. The paper discusses the implications of these regulations for machine learning and highlights opportunities for computer scientists to design fair and explainable algorithms. The GDPR replaces the 1995 Data Protection Directive and introduces stricter rules, including higher penalties for violations. It explicitly addresses the use of sensitive data and requires data controllers to prevent discriminatory effects. The regulation emphasizes the need for transparency and fairness in algorithmic decision-making, particularly in areas like credit, insurance, and social networks. The paper explores the challenges posed by the GDPR, including the need to avoid discrimination in algorithmic decision-making and to provide explanations for algorithmic decisions. It discusses the concept of "uncertainty bias," where underrepresented groups may be unfairly disadvantaged due to the algorithm's reliance on data with limited sample sizes. The paper also highlights the importance of human interpretability in algorithm design, as the GDPR requires that algorithms be transparent and fair. The GDPR also introduces a "right to explanation," requiring that users be provided with meaningful information about the logic behind algorithmic decisions. This raises questions about how to explain complex algorithms, such as neural networks, which are often considered "black boxes." The paper discusses the challenges of explaining algorithmic decisions and the need for research into methods that can quantify the influence of input variables on outputs. The paper concludes that while the GDPR presents challenges for the machine learning community, it also provides opportunities for research into fair and transparent algorithms. The regulation emphasizes the importance of ethical design and the need for collaboration between technical and philosophical resources to ensure that algorithms are not only efficient but also fair and transparent.The European Union's General Data Protection Regulation (GDPR) introduces new rules for algorithmic decision-making, including a "right to explanation." This regulation, set to take effect in 2018, restricts automated decisions that significantly affect individuals, particularly those based on user-level predictors. It also creates a right for users to request explanations for algorithmic decisions made about them. The paper discusses the implications of these regulations for machine learning and highlights opportunities for computer scientists to design fair and explainable algorithms. The GDPR replaces the 1995 Data Protection Directive and introduces stricter rules, including higher penalties for violations. It explicitly addresses the use of sensitive data and requires data controllers to prevent discriminatory effects. The regulation emphasizes the need for transparency and fairness in algorithmic decision-making, particularly in areas like credit, insurance, and social networks. The paper explores the challenges posed by the GDPR, including the need to avoid discrimination in algorithmic decision-making and to provide explanations for algorithmic decisions. It discusses the concept of "uncertainty bias," where underrepresented groups may be unfairly disadvantaged due to the algorithm's reliance on data with limited sample sizes. The paper also highlights the importance of human interpretability in algorithm design, as the GDPR requires that algorithms be transparent and fair. The GDPR also introduces a "right to explanation," requiring that users be provided with meaningful information about the logic behind algorithmic decisions. This raises questions about how to explain complex algorithms, such as neural networks, which are often considered "black boxes." The paper discusses the challenges of explaining algorithmic decisions and the need for research into methods that can quantify the influence of input variables on outputs. The paper concludes that while the GDPR presents challenges for the machine learning community, it also provides opportunities for research into fair and transparent algorithms. The regulation emphasizes the importance of ethical design and the need for collaboration between technical and philosophical resources to ensure that algorithms are not only efficient but also fair and transparent.