Study: AI-generated content reuse judged less unethical than human work

A recent study available on arXiv (2604.26956) investigates the moral judgments surrounding the reuse of content generated by AI, particularly in relation to moral patiency and ownership perceptions. In this experiment, participants assessed two manuscripts that were largely similar, with the original author identified as either a human, an AI system, or an AI entity with a human-like name. The findings indicated that the act of copying AI-generated material was viewed as less unethical, less plagiaristic, and less likely to induce guilt compared to copying work created by humans. Mediation analyses indicated that this leniency was linked to lower perceptions of AI's ability to experience harm and a heightened sense of ownership regarding the human author reusing AI content. Additionally, anthropomorphic elements influenced moral judgments by diminishing perceived ownership. These results provide insight into how individuals morally disengage when utilizing AI-generated material.