Placing Blame in Multi-agent Systems

The article discusses the concept of blame, the acts of placing blame, and blaming in the context of multi-agent systems (MAS). The conceptual resources used for analyzing MASs and their functioning mainly concentrate on successful action while attempts to understand the phenomena concerning erring, making mistakes, and wrongdoing have received less attention. Legal accountability and moral responsibility have a central role in discussions concerning the detrimental and harmful effects of MASs. However, encounters with failing systems give rise to blame and placing blame. The article analyzes the psychological, social, and moral connotations of blame and blaming and their links to responsibility and accountability. It also explores whether the ideals of responsibility and accountability should be complemented with blame-analysis to cover various failure modes and cases of non-ideal occurrence of human-machine cooperation.