An artificial neural network approach for creating an ethical artificial agent

Abstract

Autonomous robotic systems and intelligent artificial agents' capability have advanced dramatically. Since the intelligent artificial agents have been developing more autonomous and human-like, the capability of them to make moral decisions becomes an important issue. In this work we developed an artificial neutral network which considered various effective factors for ethical assessment of an action to determine that if a behavior or an action is ethically permissible or not. We integrated this net to the BDI-agent model as a part of its reasoning process to behave ethically in various environments.