PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 28, 20263 citations

Effects of AI companions’ sycophancy and emotional mimicry on consumers’ continuance intention and social wellbeing

View Full Paper
DLDaisy LeeCWCalvin WanPNPeggy M. L. Ng

Key Points

  • The study aims to explore how AI sycophancy and emotional mimicry influence consumers' wellbeing and their intention to continue using AI companions.
  • Conducted a 2 × 2 online experiment
  • Involved 636 Chinese respondents
  • Investigated moderating effects of emotional mimicry on AI sycophancy impact
  • AI companions with low sycophancy provide better social support, enhancing wellbeing and continuance intention.
  • Emotional mimicry by AI reduces the negative effects of sycophancy on emotional support but not on informational support.

Abstract

Consumers increasingly rely on social or conversational AI for companionship. While AI companions tend to be excessively agreeable with users (AI sycophancy) to boost engagement and usage, sycophantic AI may negatively impact consumers’ wellbeing. This study employed a 2 × 2 online experiment with 636 Chinese respondents to investigate the moderating effects of emotional mimicry (nonverbal expression) on the impact of AI sycophancy (verbal content) on users’ continuance intention and social wellbeing. Results indicated that AI companions with low sycophancy provide better social support to users, subsequently enhancing continuance intention and wellbeing. AI companions that mimic users’ emotions reduce the negative impact of AI sycophancy on emotional support but not informational support. Theoretical and practical implications of ethical and responsible AI companion design are discussed.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Lee et al. (2026) studied this question.

synapsesocial.com/papers/69a287350a974eb0d3c02b5ahttps://doi.org/10.1080/10447318.2026.2626809
Ask AI
Helpful
Bookmark
Share
View Full Paper