Trust in ChatGPT for political information consumption: the roles of use, perceived threat, and political ideology

Prior research suggests that -political differences shape perceptions of technology, yet limited attention has been given to how these differences affect trust in emerging technologies and how they interact with users’ behavioral and psychological responses. Addressing this gap, we investigated the roles of ChatGPT use, perceived threat of large language models (LLMs), nd political ideology in shaping trust in ChatGPT for political information consumption. Thresults show that frequent use of ChatGPT was positively associated with trust, whereas perceived threat was negatively associated with trust. Significant two-way interactions emerged between ChatGPT use and threat perception, and between threat perception and political ideology. ChatGPT use was associated with higher trust, with a stronger association observed among individuals with higher threat perception. The negative effect of threat perception on trust was also stronger among liberals than among conservatives. However, a significant three-way interaction further showed that although higher threat perception was generally associated with lower trust, it strengthened the positive relationship between ChatGPT use and trust among liberals and moderates, but not among conservatives, whose trust remained relatively stable across threat levels. These findings highlight the importance of considering socio-political identity when examining trust formation in emerging AI technologies.