Ethics of Artificial Intelligence Learning Path

Machine ethics and artificial moral agents

Machine ethics (or machine morality) is the research area focused on designing Artificial Moral Agents (AMAs)—robots or AI systems that behave morally or appear to do so. It draws on concepts from agency and moral agency to clarify what it means for an artificial system to make ethical decisions, and it explores ways to test whether AI decisions are genuinely ethical (e.g., proposals such as an “Ethical Turing Test”). Approaches also consider how such moral competence might be realized, including neuromorphic systems, whole-brain emulation, and the observation that large language models can approximate human moral judgments. A central challenge is that ethical behavior depends not only on the agent’s design but also on the data and environment that shape what it learns and inherits. Machine ethics is closely linked to AI control and value alignment: ensuring that increasingly capable systems pursue objectives compatible with human values and oversight. However, real-world deployment raises difficulties such as algorithmic biases (including racial, gender, language, political, and species biases), dominance by major tech firms, transparency and accountability gaps, and governance and regulation problems. The topic also intersects with concerns about AI welfare and moral status, threats to human dignity, liability in domains like self-driving cars, and the risk of weaponization when autonomy is used in military contexts.

0/2 modes complete

Ethical decision testing for AI systems

0/2 modes complete