Wing Yin Pang

The Mohist concept of xinshen (信身) is reinterpreted here within the framework of contemporary AI ethics. In the Mozi (墨子), the term xinshen appears exclusively in Exalting Unity III (尚同下) and is absent from all other pre-Qin texts. While it has traditionally been rendered as “integrity,” “self-confidence,” or “belief in others,” this article argues that xinshen should be understood as a dual structure: the delegation of perception and action to trusted agents, and the extension of one’s moral virtues through them. This reconceptualization reframes trust not as passive belief but as ethically guided reliance. When applied to human-AI relations, it reveals that although current AI systems substantially extend human cognition, they lack the trustworthiness and moral agency required to meet the ethical expectations embodied in xinshen. Moreover, the commercial logics underlying AI development frequently prioritize profit over public welfare, thereby contravening the Mohist principle that genuine benefit must be grounded in righteousness. Recasting xinshen as a normative ideal of ethically conditioned reliance thus offers a distinctive contribution to AI ethics: it shifts attention from individual belief or technical remediation to the moral foundations of institutional design and decision-making. Ultimately, xinshen is proposed not as a descriptive account of what AI currently is, but as an ethical orientation for what it ought to become—a framework for cultivating trustworthy, virtue-aligned, and socially responsible intelligent systems.