本研究探討在公共議題的虛擬展覽中，AI 朋友是否能促進使用者的親社會行為，並進一步分析不同 AI 外觀設計與心理機制所產生的影響。過去研究多聚焦於 AI 對個人心理福祉的影響，例如提升陪伴感、自尊或情緒支持，但較少探討這些心理效果是否能進一步轉化為實際的助人行為。因此，本研究以流浪動物虛擬展覽為情境，探討 AI 作為朋友或陪伴角色對捐贈意願的影響，以及狀態自尊與自我提升在其中所扮演的角色。

本研究以Unity 建立虛擬流浪動物展覽，並將受試者隨機分配至四種觀展情境，分別為「擬動物化 AI 朋友」、「擬人化 AI 朋友」、「真實人類朋友組」以及「獨自逛展組」。受試者完成觀展後，需填寫狀態自尊、自我提升與捐贈意願等相關量表，以分析不同陪伴情境對親社會行為的影響。

研究結果顯示，與 AI 朋友共同觀展的受試者，其流浪動物捐贈意願顯著高於獨自觀展組，代表 AI 陪伴者在公共議題情境中，可能具備促發親社會行為的潛力。其中，擬動物化 AI 相較於擬人化 AI，在親社會行為上呈現較高趨勢。質性分析結果進一步指出，擬動物化 AI 較容易降低社交壓力與被評價感，使受試者更容易投入情境並產生同理心；相較之下，擬人化 AI 則較容易引發不自然感、恐怖谷效應與社交評價焦慮。

此外，本研究原先假設「狀態自尊 → 自我提升 → 親社會行為」會形成中介效果，但研究結果並未支持此假設。進一步分析發現g，在 AI 陪伴情境中，狀態自尊的提升可正向預測親社會行為，狀態自尊提升較高的人傾向展現較多的親社會行為。相較之下，自我提升未能有效預測親社會行為，亦未形成顯著中介效果。整體而言，AI 陪伴情境中的親社會行為可能不完全來自自我提升或社會比較機制，而更可能與低社交壓力、情感安全感與情境投入有關。

整體而言，本研究認為，在公共議題的數位展覽中，AI 陪伴者不僅能提供情感支持，也可能進一步提升使用者的情感參與與親社會行為。其中，擬動物化 AI 因具備較低的社會威脅感與較高的情感親近性，可能比擬人化 AI 更適合作為公共議題中的陪伴角色。本研究結果亦可作為未來公共議題展覽、AI 角色設計與沉浸式互動體驗之參考依據。

This study examines whether AI, as a friend or companion, can promote prosocial behav-ior in a virtual stray animal exhibition, and how different AI appearance designs affect users’ psychological responses. It also explores the roles of state self-esteem and self-enhancement in this process.

A Unity-based virtual exhibition was conducted. Participants were randomly assigned to one of four conditions: a zoomorphic AI companion, a humanoid AI companion, a real human companion, and a solo viewing condition. After the exhibition, participants completed measures of state self-esteem, self-enhancement, and donation intention.

The results show that participants with an AI companion reported significantly higher do-nation intention than those in the solo condition, indicating that AI companionship may pro-mote prosocial behavior in public issue contexts. Among the AI conditions, the zoomorphic AI showed a higher tendency than the humanoid AI in promoting prosocial responses. Qualita-tive findings suggest that zoomorphic AI reduces social pressure and evaluation anxiety, while humanoid AI may trigger uncanny feelings, discomfort, and social evaluation anxiety.

The proposed mediation model (“state self-esteem → self-enhancement → prosocial be-havior”) was not supported. However, state self-esteem positively predicted prosocial behavior in AI conditions, while self-enhancement did nogt predict prosocial behavior or form a signifi-cant mediation effect. This suggests that prosocial behavior in AI contexts may be driven more by reduced social pressure, emotional safety, and situational engagement than by traditional self-enhancement or social comparison mechanisms.

Overall, AI companions can enhance emotional engagement and prosocial intentions in public issue exhibitions. Zoomorphic AI appears particularly suitable due to its lower social threat and higher emotional approachability. The findings may serve as a reference for future public issue exhibitions, AI character design, and immersive interactive experiences.

[1] Abendschein, B., Edwards, A., & Edwards, C. (2022). Novelty experience in pro-longed interaction: A qualitative study of socially-isolated college students’ in-home use of a robot companion animal. Frontiers in Robotics and AI, 9, Article 733078. https://doi.org/10.3389/frobt.2022.733078

[2] Ahmed, E., Buruk, O.‘, & Hamari, J. (2024). Human–robot companionship: Cur-rent trends and future agenda. International Journal of Social Robotics, 16, 1809–1860. https://doi.org/10.1007/s12369-024-01160-y

[3] Ho, A., Hancock, J., & Miner, A. S. (2018). Psychological, relational, and emotional effects of self-disclosure after conversations with a chatbot. Journal of Communication, 68(4), 712–733. https://doi.org/10.1093/joc/jqy026

[4] Bar-Tal, D. (1986). Altruistic motivation to help: Definition, utility, and operationalization. Humboldt Journal of Social Relations, 13(1/2), 3 – 14. http://www.jstor.org/stable/23262656

[5] Brown, J. D., Collins, R. L., & Schmidt, G. W. (1988). Self-esteem and direct versus indirect forms of self-enhancement. Journal of Personality and Social Psychology, 55(3), 445–453. https://doi.org/10.1037/0022-3514.55.3.445

[6] Barker, S. B., & Wolen, A. R. (2008). gThe benefits of human – companion animal interaction: A review. Journal of Veterinary Medical Education, 35(4), 487 – 495. https://doi.org/10.3138/jvme.35.4.487

[7] Caprara, G. V., Steca, P., Zelli, A., & Capanna, C. (2005). A new scale for measur-ing adults’ prosocialness. European Journal of Psychological Assessment, 21(2), 77–89. https://doi.org/10.1027/1015-5759.21.2.77

[8] Bartneck, C., & Forlizzi, J. (2004). A design-centred framework for social human-robot interaction. In Proceedings of the 13th IEEE International Workshop on Robot and Human Interactive Communication (RO-MAN 2004) (pp. 591 – 594). IEEE. https://doi.org/10.1109/ROMAN.2004.1374827

[9] Nass, C., Steuer, J., & Tauber, E. R. (1994). Computers are social actors. In Proceedings
of the SIGCHI Conference on Human Factors in Computing Systems (CHI ’94) (pp. 72–78). Association for Computing Machinery. https://doi.org/10.1145/191666.191703

[10] Dibble, J. L., Levine, T. R., & Park, H. S. (2012). The Unidimensional Relationship Close-ness Scale (URCS): Reliability and validity evidence for a new measure of relationship closeness. Psychological Assessment, 24(3), 565–572. https://doi.org/10.1037/a0026265

[11] Edmondson, A. (1999). Psychological safety and learning behavior in work teams. Ad-ministrative Science Quarterly, 44(2), 350–383. https://doi.org/10.2307/2666999

[12] Epley, N., & Dunning, D. (2000). Feeling ”holier than thou”: Are self-serving assess-ments produced by errors in self- or social prediction? Journal of Personality and Social Psychology, 79(6), 861–875. https://doi.org/10.1037/0022-3514.79.6.861

[13] Gur, T., & Maaravi, Y. (2025). The algorithm of friendship: Literature re-view and integrative model of relationships between humans and artificial in-telligence (AI). Behaviour & Information Technology, 44(14), 3446 – 3466. https://doi.org/10.1080/0144929X.2025.2502467

[14] Gates, M., Walker, J., Zito, S., & Dale, A. (2019). A survey of opinions towards dog and cat management policy issues in New Zealand. New Zealand Veterinary Journal, 67(6), 315–322. https://doi.org/10.1080/00480169.2019.1645627

[15] He, S., & Chen, Y. (2026). Power distancegand psychological safety in LLM counseling: Effects on self-efficacy with implications for mental health-relevant behavior change. Be-
havioral Sciences, 16(2), Article 241. https://doi.org/10.3390/bs16020241

[16] Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for mea-suring state self-esteem. Journal of Personality and Social Psychology, 60, 895–910. https://doi.org/10.1037/0022-3514.60.6.895

[17] Homans, G. C. (1958). Social behavior as exchange. American Journal of Sociology, 63(6), 597–606. https://doi.org/10.1086/222355

[18] Huang, G., & Wang, S. (2023). Is artificial intelligence more persuasive than humans? A meta-analysis. Journal of Communication, 73(6), 552 – 562.
https://doi.org/10.1093/joc/jqad024

[19] Stein, J.-P., & Ohler, P. (2017). Venturing into the uncanny valley of mind—The influence of mind attribution on the acceptance of human-like characters in a virtual reality setting.
Cognition, 160, 43–50. https://doi.org/10.1016/j.cognition.2016.12.010

[20] Arnold, L. (2016). Emobie™: A robot companion for children with anxiety. In Proceed-ings of the 2016 11th ACM/IEEE International Conference on Human-Robot Interaction (HRI) (pp. 413–414). IEEE. https://doi.org/10.1109/HRI.2016.7451782

[21] Leitner, J. B., & Forbes, C. E. (2015). The role of implicit mechanisms in buffering self-esteem from social threats. In Z. Jin (Ed.), Exploring implicit cognition: Learning, memory, and social cognitive processes (pp. 183–204). IGI Global Scientific Publishing. https://doi.org/10.4018/978-1-4666-6599-6.ch009

[22] Mori, M., MacDorman, K. F., & Kageki, N. (2012). The uncanny valley [From the field]. IEEE Robotics & Automation Magazine, 19(2), 98 – 100. https://doi.org/10.1109/MRA.2012.2192811

[23] Munir, S. M. I., Mokhtar, M. I., & Arham, A. F. (2023). Public perspectives on strays and companion animal management in Malaysia. BMC Public Health, 23, Article 1428. https://doi.org/10.1186/s12889-023-16276-5

[24] Ariya, P., Khanchai, S., Intawong, K., & Puritat, K. (2025). Enhancing tex-tile heritage engagement through generative AI-based virtual assistants in vir-
tual reality museums. Computers & Education: X Reality, 7, Article 100112. https://doi.org/10.1016/j.cexr.2025.100112g

[25] Brandtzaeg, P. B., Skjuve, M., & Følstad, A. (2022). My AI friend: How users of a social chatbot understand their human–AI friendship. Human Communication Research, 48(3), 404–429. https://doi.org/10.1093/hcr/hqac008

[26] Rogers, C. R. (1957). The necessary and sufficient conditions of therapeu-tic personality change. Journal of Consulting Psychology, 21(2), 95 – 103. https://doi.org/10.1037/h0045357

[27] Rosenberg, M. (1965). Rosenberg Self-Esteem Scale (RSES). APA PsycTests. https://doi.org/10.1037/t01038-000

[28] Rosenberg, M., Schooler, C., Schoenbach, C., & Rosenberg, F. (1995). Global self-esteem and specific self-esteem: Different concepts, different outcomes. American Sociological Review, 60(1), 141–156. https://doi.org/10.2307/2096350

[29] Salah, M., Alhalbusi, H., & Ismail, M. M., et al. (2024). Chatting with ChatGPT: De-coding the mind of chatbot users and unveiling the intricate connections between user perception, trust and stereotype perception on self-esteem and psychological well-being.
Current Psychology, 43, 7843–7858. https://doi.org/10.1007/s12144-023-04989-0

[30] Sedikides, C., & Gregg, A. P. (2008). Self-enhancement: Food for thought. Per-spectives on Psychological Science, 3(2), 102 – 116. https://doi.org/10.1111/j.1745-6916.2008.00068.x

[31] Pfattheicher, S., Nielsen, Y. A., & Thielmann, I. (2022). Prosocial behavior and altru-ism: A review of concepts and definitions. Current Opinion in Psychology, 44, 124–129. https://doi.org/10.1016/j.copsyc.2021.08.021

[32] Araujo, T. (2018). Living up to the chatbot hype: The influence of anthropo-morphic design cues and communicative agency framing on conversational agent and company perceptions. Computers in Human Behavior, 85, 183 – 189. https://doi.org/10.1016/j.chb.2018.03.051

[33] Tavoli, A., Melyani, M., & Bakhtiari, M., et al. (2009). The Brief Fear of Negative Eval-
uation Scale (BFNE): Translation and validation study of the Iranian version. BMC Psy-chiatry, 9, Article 42. https://doi.org/10.118g6/1471-244X-9-42

[34] Watson, D., & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 33(4), 448–457. https://doi.org/10.1037/h0027806

[35] Yao, X., & Xi, Y. (2025). From assistants to digital beings: Exploring an-thropomorphism, humanness perception, and AI anxiety in large-language-model chatbots. Social Science Computer Review. Advance online publication. https://doi.org/10.1177/08944393251354976

[36] Seo, J. Y., & Scammon, D. L. (2014). Does feeling holier than others predict good deeds? Self-construal, self-enhancement and helping behavior. Journal of Consumer Marketing, 31(6/7), 441–451. https://doi.org/10.1108/JCM-06-2014-1029

[37] Zhao, Y., Qian, W., Chen, Y., Wu, D., Luo, Y., Gao, C., Wu, K., & Liu, Z.
(2025). Effect of an AI agent trained on a large language model (LLM) as an inter-vention for depression and anxiety symptoms in young adults: A 28-day randomized controlled trial. Applied Psychology: Health and Well-Being, 17(5), Article e70067. https://doi.org/10.1111/aphw.70067

[38] Zhu, Z., Tan, Y., Yamashita, N., Lee, Y.-C., & Zhang, R. (2025). The benefits of proso-ciality towards AI agents: Examining the effects of helping AI agents on human well-being. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (CHI ’25) (Article 722, pp. 1 – 18). Association for Computing Machinery. https://doi.org/10.1145/3706598.3713116