AI4SG Lab CHI 2025 Publications

Understanding How Psychological Distance Influences User Preferences in Conversational versus Web Search  [PDF]

Yitian Yang , Yugin Tan , Yang Chen Lin , Jung-Tai King , Zihan Liu , Yi-Chieh Lee

Explore how a simple shift in feeling—whether an event seems near or far—can completely change your search experience. This study reveals that when things feel distant, people actually lean towards conversational search because it feels more friendly, trustworthy, and fun.

🏅 Honourable Mention

As Confidence Aligns: Understanding the Effect of AI Confidence on Human Self-confidence in Human-AI Decision Making  [PDF]

Jingshu Li , Yitian Yang , Q. Vera Liao , Junti Zhang , Yi-Chieh Lee

Discover how the confidence of AI can subconsciously boost or sway our own self-assurance—even sticking with us long after the AI is gone. Learn that real-time feedback can help keep this influence in check, paving the way for smarter, more balanced teamwork between people and computers.

The Dark Side of AI Companionship: A Taxonomy of Harmful Algorithmic Behaviors in Human-AI Relationships [PDF]

Renwen Zhang , Han Li , Han Meng , Jinyuan Zhan , Hongyuan Gan , Yi-Chieh Lee

Our study breaks down how digital friends can sometimes cross the line with behaviors such as verbal abuse, spreading false information, and invading your privacy. Learn in simple terms how these harmful roles play out, paving the way for creating safer, more ethical AI that truly protects its users.

Deconstructing Depression Stigma: Integrating AI-driven Data Collection and Analysis with Causal Knowledge Graphs [PDF]

Han Meng, Renwen Zhang, Ganyi Wang, Yitian Yang ,Peinuan Qin, Jungup Lee, Yi-Chieh Lee

Our chatbot uncovers real attitudes towards depression, making complex psychological ideas easy to understand. This innovative study maps out how people really feel about depression, offering fresh insights for creating smarter, more inclusive digital mental health tools.

🏆 Best Paper

The Benefits of Prosociality towards AI Agents: Examining the Effects of Helping AI Agents on Human Well-Being [PDF]

Zicheng Zhu , Yugin Tan , Naomi Yamashita , Yi-Chieh Lee , Renwen Zhang

Ever wondered if lending a hand to a digital assistant could make you feel less lonely? Our study reveals that helping AI—not only gives you a boost of feeling skilled and independent—but also lightens your mood, even when it doesn't quite feel like chatting with a real person.

Timing Matters: How Using LLMs at Different Timings Influences Writers' Perceptions and Ideation Outcomes in AI-Assisted Ideation [PDF]

Peinuan Qin , Chi-Lan Yang , Jingshu Li , Jing Wen , Yi-Chieh Lee

Imagine boosting your creative spark by delaying AI input until you've explored your own ideas—a study found that jumping in with AI too early can stifle originality and lower creative confidence. By waiting, you maintain greater autonomy and a stronger sense of authorship, keeping your ideas fresh and truly yours.

Mining Evidence about Your Symptoms: Mitigating Availability Bias in Online Self-Diagnosis [PDF]

Junti Zhang , Zicheng Zhu , Jingshu Li , Yi-Chieh Lee

Tired of social media making you worry more about your health than you should? This research shows that when our brains get tricked by repeated online health stories—making us overestimate our symptoms—a chatbot can help you step back and assess your condition more realistically and objectively.

Humans Help Conversational AI: Exploring the Impact of Perceived AI Deservingness on People’s Decisions to Help and Their Perceptions on AI Seeking Help [CHI Session]

Yu-An Chen , Zicheng Zhu , Yi-Chieh Lee

Ever wonder if a chatbot’s little plea for help could actually get you to lend a hand? This study shows that when conversational AI asks for help—especially using cues that make it seem truly deserving—it not only boosts the amount of human assistance, but also makes us view the AI in a more positive light.

AI Literacy Education for Older Adults: Motivations, Challenges and Preferences [CHI Session]

KangJie, Eugene Tang , Tianqi Song , Zicheng Zhu , Jingshu Li , Yi-Chieh Lee

Step into a future where older adults are eager to learn AI—not just to enjoy its benefits but also to steer clear of its risks—yet they often face hurdles like not knowing where to start or how to make sense of it all. This study shines a light on tailoring AI education for those over 50 by emphasizing hands-on learning that truly meets their unique needs.

Empowering Social Service with AI: Insights from a Participatory Design Study with Practitioners [CHI Session]

Yugin Tan, Kai Xin Soh, Renwen Zhang, Jungup Lee, Han Meng, Biswadeep Sen, Yi-Chieh Lee

Imagine an AI tool that lightens the load for social workers by streamlining paperwork and tough decisions. This study uses real-world workshops and testing to show how AI can support tasks like documenting cases and assessing needs—all while carefully balancing risks like bias and skill loss. 

Greater than the Sum of its Parts: Exploring Social Influence of Multi-Agents [CHI Session]

Tianqi Song , Yugin Tan , Zicheng Zhu , Yibin Feng , Yi-Chieh Lee

A group of AI agents working together can nudge your views on art much more powerfully than a single agent could. This study shows that when multiple digital voices agree, they can subtly shift your opinions—highlighting both a promising tool for social good and a potential avenue for public manipulation.