不诚实
心理学
人际交往
诚实
社会心理学
生成语法
透明度(行为)
工艺
自我表露
生成模型
人际关系
机制(生物学)
人格
互联网隐私
作者
Danielle Hass,Colleen Kirk,Julian Givi,Danielle Hass,Colleen Kirk,Julian Givi
摘要
ABSTRACT The rise of generative AI (GenAI), such as ChatGPT, has transformed user communication. Consumers now use GenAI to craft personal messages, including birthday wishes, congratulatory notes, and even love poems. Although GenAI's technical capabilities are well documented, less is known about its psychological impacts in interpersonal contexts. This research investigates how consumers respond when using GenAI to author emotionally laden messages. Drawing on appraisal theory, we propose and find that GenAI use elicits guilt when presenting AI‐generated messages as one's own. Across five preregistered experiments spanning varied messaging contexts, we demonstrate that consumers feel more guilt when they use GenAI for emotionally laden messages (vs. composing the message themselves). This effect is driven by perceived dishonesty. Supporting our dishonesty framework, we find that guilt also arises when a message is secretly written by a friend, but not when it comes on a pre‐printed greeting card. Further consistent with a dishonesty account, the effect weakens when the message is written but not delivered, and when the recipient is socially distant (vs. close). This work advances theory by identifying perceived dishonesty as a key psychological mechanism behind consumer discomfort in GenAI‐assisted personal communications and by incorporating appraisal theory as an explanatory framework. Practically, our findings suggest that firms and developers should design GenAI tools that emphasize transparency and co‐creation to reduce negative emotional responses, especially in emotionally significant situations where honesty matters.
科研通智能强力驱动
Strongly Powered by AbleSci AI