AI Ethics is a course designed for Master 1 Mathematics students that explores the ethical, mathematical, and societal dimensions of Artificial Intelligence and automated decision-making systems. This course examines how mathematical models, statistical methods, optimization techniques, and data-driven algorithms are used to make predictions and decisions that can affect individuals and society. Particular attention is given to the relationship between mathematical performance and ethical responsibility.
The course addresses the role of probability, statistics, uncertainty, risk, and decision theory in AI systems. Therefore, students will study the main sources of bias in data and algorithms, methods for evaluating algorithmic fairness, and the mathematical foundations of different fairness criteria. The course also introduces transparency and explainability, including the interpretation of mathematical models. Then, students will also be introduced to fundamental concepts related to privacy, data protection, accountability, human oversight, and AI governance.
Through mathematical examples, practical case studies, and critical analysis of real-world applications such as loan approval, recruitment, healthcare, education, and automated decision-making, students will learn to identify ethical risks and evaluate AI systems from both mathematical and societal perspectives.
The course aims to develop a critical understanding of the fact that a model can be mathematically accurate while still producing biased, unfair, or socially undesirable outcomes. Students will therefore learn to consider not only how a model works, but also what it optimizes, what assumptions it makes, who may be affected, and whether its decisions are fair and responsible.