Calculation of the Categorical Imperative in a Multi-Agent Environment

With the growing need to represent human ethical knowledge in the field of artificial intelligence, and considering the heavy legacy of philosophical and doctrinal issues specific to various schools of thought, the major challenge lies in representing this knowledge computationally. In a multi-agent scenario, if one wishes to integrate Kantian ethical deontology, which requires a formulation consistent with the theoretical definitions represented by the categorical imperative, this consistency depends on two fundamental conditions: generalizability without contradiction, and adherence to universal moral principles. Here, we propose a simple and detailed method for calculating this consistency, by combining logical and mathematical approaches.