Automated conflict resolution: Sharing sequentially triggered losses

Abstract

When actions by one agent force another to deviate from their agreements with a third, “victim” turns into “injurer” in the chain’s subsequent steps. Should the chain’s initiator be responsible only for the direct harm they cause or also bear some of the indirect losses they trigger? Through an axiomatic approach, we characterize the class of fixed-fraction rules, which strike a balance between incentives for accident prevention on the one hand and fairness in terms of how liabilities are assigned on the other. Their simple design make the rules ideal for practical implementation through smart contracts, enabling automated conflict resolution.