The hedonic model of Kant's categorical imperative

To represent an ethics of philosophical origin in computational form, thereby enabling agents to adhere to universal and unconditional ethical rules, we propose modifying the hedonic model by introducing ethical constraints based on rule generalization. This enhancement aims to enrich interactions within a heterogeneous multi-agent environment. We specifically explore integrating Immanuel Kant’s ethical principles, particularly the categorical imperative, into a multi-agent hedonic game model. While traditionally centered on agents' personal preferences, this model can be refined by incorporating strict moral imperatives. The outcome of this approach is a formal representation of ethical principles, providing a framework for ethical decision-making in multi-agent systems.