AI as Moral Actor

This line of research examines AI as a moral actor in public communication, asking when people accept AI as a legitimate voice for correcting wrongdoing and enforcing norms. Particularly, we study how people evaluate and behaviorally respond to AI when it speaks against harmful behavior, what psychological mechanisms shape those judgments and reactions, and how design choices influence acceptance of and endorsement of AI as a visible community participant. This work helps clarify when AI can support healthier online communities through scalable interventions against toxicity and when human moral authority remains essential.

Selected Work:

Zhao, P., Bazarova, N. N., & Margolin, D. (2026). Involving AI in public moral dialogue: Can AI serve as a legitimate moral actor to address problematic online speech? Available at SSRN.

Zhao, P. (Manuscript in preparation). AI as a community participant: Effects on behavioral norm expression in online communities. 

AI for Social Support

This line of work investigates how AI can expand access to social support while preserving the unique value of human care. Rather than asking whether AI can replace human support, we examine how humans and AI can work together by identifying the support functions each is best suited to provide. This work aims to inform the design of AI systems that make supportive communication more effective, accessible, and human-centered.

Selected Work:

Zhao, P., Li, H, & Park, S. (Manuscript in preparation). Leverage AI to optimize social support. 

AI Companion: Impact and Interventions

This line of research examines how AI companions are becoming part of people’s social and emotional lives. We investigate how these systems shape expectations for human relationships, influence reliance on AI for connection, and affect how people manage stress and distress. We also develop and test AI literacy and intervention approaches that support more reflective AI companion use and help people maintain healthy relationships with both AI systems and other people.

Online Toxicity Interventions

This line of research examines how interventions can promote prosocial norms in online communities rife with problematic content. We focus on justice-oriented counterspeech interventions, investigating when, how, and why they receive community support and help foster healthier digital spaces. Through experiments on simulated social media platforms, this work informs evidence-based approaches to online toxicity intervention.

Selected Work:

Zhao, P., Bazarova, N. N., Bae, I., Hui, W., Kizilcec, R., & Margolin, D. (2025). Restorative justice appeals trump retributive vigilance on social media. PNAS Nexus, 4, pgaf255. 

Zhao, P., Bazarova, N. N., DiFranzo, D., Hui, W., Kizilcec, R., & Margolin, D. (2024). Standing up to problematic content on social media: Which objection strategies draw the audience’s approval? Journal of Computer-Mediated Communication, 29(1), 1-13.