Paper Type
Complete
Paper Number
1252
Description
Given the proliferation of AI-driven social bots, social media platforms face the dilemma of whether disclosing bot identity can enhance transparency without undermining user engagement. Previous research on AI identity disclosure in commercial settings shows mixed effects due to varying levels of trust and AI aversion. Considering the complexity of engagement and motivations on social media, past findings may not directly migrate to our understanding of potential disclosure effects in social contexts. Based on social identity theory, this study explores how bot identity disclosure influences human high-effort (e.g., commenting) and low-effort (e.g., liking, retweeting) engagement behaviors. Employing a natural experiment informed by Twitter’s #GoodBot policy, DID models reveal that bot identity disclosure negatively affects high-effort engagement but not low-effort engagement, with these effects magnified by bots’ social influence. This research extends the understanding of identity disclosure and offers practical insights for social platforms.
Recommended Citation
Zhang, Yiqun; Wei, Zhen; Song, Danyang; and Chen, Xi, "Social Bot Identity Disclosure and User Engagement: Evidence from a Natural Experiment on Twitter" (2024). PACIS 2024 Proceedings. 11.
https://aisel.aisnet.org/pacis2024/track13_hcinteract/track13_hcinteract/11
Social Bot Identity Disclosure and User Engagement: Evidence from a Natural Experiment on Twitter
Given the proliferation of AI-driven social bots, social media platforms face the dilemma of whether disclosing bot identity can enhance transparency without undermining user engagement. Previous research on AI identity disclosure in commercial settings shows mixed effects due to varying levels of trust and AI aversion. Considering the complexity of engagement and motivations on social media, past findings may not directly migrate to our understanding of potential disclosure effects in social contexts. Based on social identity theory, this study explores how bot identity disclosure influences human high-effort (e.g., commenting) and low-effort (e.g., liking, retweeting) engagement behaviors. Employing a natural experiment informed by Twitter’s #GoodBot policy, DID models reveal that bot identity disclosure negatively affects high-effort engagement but not low-effort engagement, with these effects magnified by bots’ social influence. This research extends the understanding of identity disclosure and offers practical insights for social platforms.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.
Comments
Interaction