How Does Deep Synthesis Content Erode Trust in Social Media? — An Analysis from the Dimensions of Platform, Information, and Users Based on a Survey Experiment

Xinhui Wu; Sheng Lin

doi:10.54254/2753-7048/2025.26637

Lecture Notes in Education Psychology and Public MediaOpen access

How Does Deep Synthesis Content Erode Trust in Social Media? — An Analysis from the Dimensions of Platform, Information, and Users Based on a Survey Experiment

Research Article

Open Access

How Does Deep Synthesis Content Erode Trust in Social Media? — An Analysis from the Dimensions of Platform, Information, and Users Based on a Survey Experiment

Xinhui Wu ¹, Sheng Lin ^2*

¹ Hangzhou Dianzi University

² Hangzhou Dianzi University

^*Corresponding author: l1332286868@163.com

Published on 9 September 2025

LNEP Vol.117

ISSN (Print): 2753-7056

ISSN (Online): 2753-7048

ISBN (Print): 978-1-80590-361-1

ISBN (Online): 978-1-80590-362-8

Download Cover

Abstract

Deep-synthesis technology—an artificial intelligence technique used primarily for audiovisual content—has been misused in ways that are systematically deconstructing the trust ecosystem of social media. This study innovatively unpacks social media trust into three dimensions (platform trust, information trust, and user trust) and, using a survey-experiment design (n = 522), examines the differentiated erosive effects of deep-synthesis content within social media platforms. Taking Douyin (the Chinese version of TikTok) as the context, we employ five categories of deep-synthesis videos as stimuli and control for topic-related confounds. The findings show that deep-synthesis content exerts a negative impact on social media trust, with information trust being the most severely damaged, followed by platform trust; user/interpersonal trust exhibits a lagging and comparatively weak effect. These results reveal differentiated pathways through which technological alienation disrupts trust mechanisms and provide a theoretical basis for platform governance and user-level cognitive interventions.

Keywords:

deep-synthesis technology, social media trust, survey experiment

View PDF

References

[1]. Twomey, J., Ching, D., Aylett, M. P., et al. (2023). Do deepfake videos undermine our epistemic trust? A thematic analysis of tweets that discuss deepfakes in the Russian invasion of Ukraine. PLOS ONE, 18(10), e291668.

[2]. Mirsky, Y., & Lee, W. (2022). The creation and detection of deepfakes. ACM Computing Surveys, 54(1), 1–41.

[3]. Qi An Xin Group. (2024). 2024 Artificial Intelligence Security Report [Report]. Qi An Xin Group.

[4]. Li, J., Wen, J., Xu, Q., et al. (2024). Continued use or negative use: A study on dynamic user behavior of social media from the perspective of ambivalent attitudes. Information Science, 42(9), 100–111.

[5]. Li, Y., Zhang, X., & Wang, S. (2018). Research on the quality of health information in social media: An analysis based on features of true and false health information. Journal of the China Society for Scientific and Technical Information, 37(3), 294–304.

[6]. Yu, M., Yu, S., & Liu, R. (2024). Healthcare workers’ correction intentions of false health information on social media: Based on SEM and fsQCA methods. Journal of Information Resources Management, 14(3), 104–120.

[7]. Peng, L., & Jiang, X. (2022). The generative mechanism of health information avoidance behavior of social media users from the perspective of risk perception. Library and Information Service, 66(22), 55–65.

[8]. Gallotti, R., Valle, F., Castaldo, N., et al. (2020). Assessing the risks of “infodemics” in response to COVID-19 epidemics. Nature Human Behaviour, 4(12), 1285–1293.

[9]. Chu, Y., Wang, P., & Hu, S. (2025). The mechanism of social media health information user trust formation. Modern Information, 45(1), 97–111.

[10]. Zhang, A., & Wang, F. (2021). Deepfakes and the mutation of political opinion from the perspective of artificial intelligence. Journal of Hohai University (Philosophy and Social Sciences), 23(4), 29–36.

[11]. Karasavva, V., & NoorbhaI, A. (2021). The real threat of deepfake pornography: A review of Canadian policy. Cyberpsychology, Behavior, and Social Networking, 24(3), 203–209.

[12]. Carnevale, A., Delgado, C. F., & Bisconti, P. (2023). Hybrid ethics for generative AI: Some philosophical inquiries on GANs. Humana. Mente Journal of Philosophical Studies, 16(44), 33–56.

[13]. Leyva, R., & Beckett, C. (2020). Testing and unpacking the effects of digital fake news: On presidential candidate evaluations and voter support. AI & Society, 35, 969–980.

[14]. Zhang, D. (2023). The impact and governance of deepfake technology on journalism. Youth Journalist, (23), 41–45.

[15]. Zhou, Z. (2024). Rediscovering the state: An attempt to expand the concept of media trust in China. Chinese Journal of Journalism & Communication, 46(7), 28–53.

[16]. Dwyer, C., Hiltz, S., & Passerini, K. (2007). Trust and privacy concern within social networking sites: A comparison of Facebook and MySpace. AMCIS 2007 Proceedings, 339.

[17]. Niu, J., & Meng, X. (2019). The influence of social media trust on privacy risk perception and self-disclosure: The mediating effect of online interpersonal trust. Chinese Journal of Journalism & Communication, 41(7), 91–109.

[18]. Engelke, K. M., Hase, V., & Wintterlin, F. (2019). On measuring trust and distrust in journalism: Reflection of the status quo and suggestions for the road ahead. Journal of Trust Research, 9(1), 66–86.

[19]. Yao, Q., Hou, M., Fu, M., et al. (2022). The impact of bullet comments on trust in mainstream media: A user–media matching perspective. Psychological Science, 45(2), 462–469.

[20]. Wang, S., Li, Z., Chen, Y., et al. (2022). Application of survey experiments in sociology: A methodological review. Sociological Review of China, 10(6), 230–252.