Abstract | ||
---|---|---|
Recent years have yielded many discussions on how to endow autonomous agents with the ability to make ethical decisions, and the need for explicit ethical reasoning and transparency is a persistent theme in this literature. We present a modular and transparent approach to equip autonomous agents with the ability to comply with ethical prescriptions, while still enacting pre-learned optimal behaviour. Our approach relies on a normative supervisor module, that integrates a theorem prover for defeasible deontic logic within the control loop of a reinforcement learning agent. The supervisor operates as both an event recorder and an on-the-fly compliance checker w.r.t. an external norm base. We successfully evaluated our approach with several tests using variations of the game Pac-Man, subject to a variety of “ethical” constraints. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1007/s10676-022-09665-8 | Ethics and Information Technology |
Keywords | DocType | Volume |
Deontic defeasible logic, Reinforcement learning, Normative reasoning, Ethical artificial intelligence | Journal | 24 |
Issue | ISSN | Citations |
4 | 1388-1957 | 0 |
PageRank | References | Authors |
0.34 | 12 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Neufeld Emery A. | 1 | 0 | 0.34 |
Ezio Bartocci | 2 | 733 | 57.55 |
Agata Ciabattoni | 3 | 510 | 59.97 |
Governatori Guido | 4 | 0 | 0.34 |