Graduate Student Research Focus: Jiacheng Huang, Ph.D. Candidate

What inspired you to pick your topic of research?

My interest in this topic grew out of two experiences that came together at the right time. First, I noticed friends and former colleagues in the communication industry beginning to use generative AI tools to draft social media content, and being genuinely surprised by the quality. That made me curious: was AI-generated content actually more engaging than what professionals were producing, or did it just feel that way?

Second, as tools like GPT-4 and their APIs became widely accessible, I realized there was an opportunity to study this question rigorously rather than anecdotally. Most of the conversation around AI in strategic communication at that point was either speculative or focused on efficiency gains. Very few studies had directly compared AI-generated and human-written messages on real engagement outcomes with controlled experiments. I wanted to go beyond the question of whether AI could write social media posts and ask whether audiences actually responded to them differently, and if so, why. That second question is what ultimately led to the development of this paper.

What contribution do you hope your research makes to the field?

This study offers both a theoretical framework and empirical evidence at a time when the field is grappling with what AI means for strategic communication practice. On the empirical side, we found that consumers rated AI-generated social media posts as more engaging than original human-written posts and that even trained, financially incentivized communication professionals did not fully close that gap. That finding challenges assumptions about the irreplaceability of human creativity in routine content production.

On the theoretical side, the study introduces the FIIT model, which identifies four message-level characteristics — fluency, interactivity, information and tone — as mechanisms through which social media engagement improves. I hope the FIIT framework gives researchers a concrete, testable vocabulary for studying AI-generated communication rather than treating AI output as a black box. For practitioners, the findings suggest that AI is most valuable as a collaborative tool: it can elevate content quality on measurable dimensions, but strategy, ethical judgment, audience understanding and brand voice still require human oversight. 

More broadly, I hope this work encourages the field to move past the "will AI replace us" debate and toward more productive questions about how communication professionals can integrate these tools responsibly and effectively.

What support did you receive from the School to be able to do this research?

This project would not have been possible without the support I received at the Hubbard School. The Hubbard School also fostered an environment where conversations with faculty and fellow graduate students in seminars, lab meetings and informal discussions helped me refine both the theoretical framing and practical implications. On the financial side, this project received funding from the Wells Fund for the Scientific Study of Strategic Communication, which covered participant recruitment and research expenses across the three studies. That support made it possible to run the kind of multi-study design the research questions demanded.