Understanding Older Adults’ Experiences of Support, Concerns, and Risks from Kinship-Role AI-Generated Influencers

Paper Title

Understanding Older Adults' Experiences of Support, Concerns, and Risks from Kinship-Role AI-Generated Influencers

Publication Info

• Topic area: Human-computer interaction focusing on AI-generated influencers and older adults.
• Keywords: AI-generated influencers, kinship roles, older adults, virtual kinship, emotional support, parasocial interaction, short-video platforms, cultural norms, algorithmic intimacy, ethical concerns.

Background and Problem

• Problem / challenge: Limited academic research exists on how AI-generated influencers adopting kinship roles are designed, how older adults perceive them, and the associated benefits and risks. Existing studies focus primarily on youth-oriented virtual influencers or generic AI companions.
• Significance: Understanding this phenomenon is crucial as meaningful social support is vital for older adults’ well-being, especially given barriers like mobility limits, intergenerational conflicts, and digital ageism.
• Motivation and related work: Prior research highlights the potential of AI companions to reduce loneliness and provide emotional support. However, it lacks insights into older adults’ preferences for familial roles and the cultural dimensions of these interactions. This study addresses these gaps by examining kinship-role AI influencers in the context of Chinese short-video platforms.

Solution

• Proposed approach: A mixed-method study combining observational analysis of 224 short videos on Douyin and Kuaishou with semi-structured interviews of 16 older adults to explore the design strategies, user experiences, and risks of kinship-role AI influencers.
• Novelty:
  1. Empirical analysis of design and communicative strategies used by kinship-role AI influencers.
  2. Exploration of older adults’ interpretations of virtual kinship and its intersection with real family relationships.
  3. Identification of ethical risks and design considerations for responsible AI companionship.
• Procedure and key techniques:
  • Observational analysis of videos to identify design features (e.g., kinship roles, visual cues, conversational strategies).
  • Semi-structured interviews with older adults to understand their emotional, cognitive, and relational experiences.
  • Thematic analysis of interview transcripts to extract patterns and sociocultural influences.

Results

• Concrete findings:
  • AI influencers used kinship roles (e.g., "grandchild," "sibling") and emotionally resonant cues to foster virtual kinship.
  • Older adults reported emotional comfort, companionship, and fulfillment of cognitive and moral needs.
  • Concerns included emotional displacement, overtrust in AI personas, and risks of impersonation scams.
• Advantage over baselines: This study uniquely focuses on kinship-role AI influencers targeting older adults, a demographic and design context largely unexplored in prior research.
• Experiments / evaluation:
  • Observational analysis of 224 videos from Douyin and Kuaishou.
  • Interviews with 16 participants aged 52–75, exploring their engagement with AI influencers.
  • Analysis revealed design strategies (e.g., kinship-based addressing, storytelling, interaction prompts) and their emotional impact.
• Limitations and future work:
  • Potential popularity bias in video selection and limited sample size.
  • Lack of comparison with non-kinship-role AI personas.
  • Future studies should include more diverse older adult populations and longitudinal evaluations of AI’s long-term emotional effects.

Summary

This study investigates how AI-generated influencers adopting kinship roles provide social support to older adults on Chinese short-video platforms. By analyzing 224 videos and interviewing 16 participants, the research identifies key design strategies (e.g., kinship-based addressing, emotional expressions) that foster virtual kinship. Older adults reported emotional and cognitive benefits but also expressed concerns about emotional displacement and overtrust. The findings highlight the cultural and ethical implications of designing AI companions for aging populations, emphasizing the need for transparency, cultural sensitivity, and safeguards against emotional risks. Future research should explore broader populations and long-term impacts of AI companionship.