The Influence of Generative Artificial Intelligence External Features on Trust and Willingness to Use

本研究聚焦生成式人工智能，以ChatGPT和文心一言为例，探讨头像类型、面孔吸引力、性别一致性等外部特征对用户信任度与使用意愿的影响机制。结合社会存在理论、性别认同一致性理论及技术接受模型分析发现，头像类型在不同场景影响各异，面孔吸引力通过“美即好”效应起作用，性别一致性可缩小心理距离促进信任，且吸引力可能放大性别一致性效应。研究为优化生成式人工智能设计提供理论参考，也指出情境多样性不足等局限及未来研究方向。This study focuses on Generative Artificial Intelligence, taking ChatGPT and Wenxin Yiyan as examples, to explore the influence mechanisms of external features, such as avatar type, facial attractiveness, and gender congruence on users’ trust and willingness to use. Analyses based on the Social Presence Theory, Gender Identity Congruence Theory, and Technology Acceptance Model reveal that avatar types have varying effects in different scenarios, facial attractiveness works through the “what is beautiful is good” effect, gender congruence can narrow psychological distance to promote trust, and attractiveness may amplify the gender congruence effect. The study provides theoretical references for optimizing Generative Artificial Intelligence design, while also pointing out limitations, such as insufficient situational diversity and future research directions.