Human Attribution of Causality to AI Across Agency, Misuse, and Misalignment

arXiv:2603.13236v1 Announce Type: new Abstract: AIrelated incidents are becoming increasingly frequent and severe, ranging from safety failures to misuse by malicious actors.