AI has attracted significant attention in the last year, initially due to the release of ChatGPT, followed by backlash and efforts at creating effective regulation. Questions of ethics underlie every aspect of AI, beginning with the question of whether it is even coherent to speak of an intelligence other than humans. This course presents current ethical issues in the development and application of artificial intelligence through a series of recent case studies. We will spend the first part of the course studying major ethical frameworks (consequentialism, deontology, virtue ethics) and closely-linked research areas within AI and machine learning. In the second part of the course, we will apply these principles to case studies from major areas of debate in AI, with a focus on the translation of ethical principles into practical decisions.The first examples from AI we will cover are existential risks in the context of utilitarianism, the "hidden" labour force of AI in the context of deontology, and the problem of replacing humans in the context of virtue ethics. For the case studies, we will first study fairness and bias in the training and deployment of machine learning models. We will ask what it means for an AI system to be "fair", and how to regulate models which are not interpretable. This is followed by the problems of copyright and large scale training datasets for generative AI models, where we will ask what constitutes unfair use of existing material when it is only being used to train. We continue in a more hypothetical lens with a discussion of whether or not an AI system could be a moral agent or patient, and what rights a non-human intelligence might have. Finally, we conclude with the alignment problem, where we focus on the practical challenges of value alignment and the plausibility of finding a set of values which could be universally accepted. In the last week of the course, students apply their learnings with group presentations on published academic research, unpacking the ethical questions underlying technical developments
4-5 units · Letter (ABCD/NP) · GER: WAY-ER
AI has attracted significant attention in the last year, initially due to the release of ChatGPT, followed by backlash and efforts at creating effective regulation. Questions of ethics underlie every aspect of AI, beginning with the question of whether it is even coherent to speak of an intelligence other than humans. This course presents current ethical issues in the development and application of artificial intelligence through a series of recent case studies. We will spend the first part of the course studying major ethical frameworks (consequentialism, deontology, virtue ethics) and closely-linked research areas within AI and machine learning. In the second part of the course, we will apply these principles to case studies from major areas of debate in AI, with a focus on the translation of ethical principles into practical decisions.The first examples from AI we will cover are existential risks in the context of utilitarianism, the "hidden" labour force of AI in the context of deontology, and the problem of replacing humans in the context of virtue ethics. For the case studies, we will first study fairness and bias in the training and deployment of machine learning models. We will ask what it means for an AI system to be "fair", and how to regulate models which are not interpretable. This is followed by the problems of copyright and large scale training datasets for generative AI models, where we will ask what constitutes unfair use of existing material when it is only being used to train. We continue in a more hypothetical lens with a discussion of whether or not an AI system could be a moral agent or patient, and what rights a non-human intelligence might have. Finally, we conclude with the alignment problem, where we focus on the practical challenges of value alignment and the plausibility of finding a set of values which could be universally accepted. In the last week of the course, students apply their learnings with group presentations on published academic research, unpacking the ethical questions underlying technical developments
Offered in Winter 2026 at Stanford University.