Welcome to the forefront of conversational AI as we explore the fascinating world of AI chatbots in our dedicated blog series. Discover the latest advancements, applications, and strategies that propel the evolution of chatbot technology. From enhancing customer interactions to streamlining business processes, these articles delve into the innovative ways artificial intelligence is shaping the landscape of automated conversational agents. Whether you’re a business owner, developer, or simply intrigued by the future of interactive technology, join us on this journey to unravel the transformative power and endless possibilities of AI chatbots.
UNC Hussman hosts Converge-Con 2025 from October 22-24. Join us for an exploration of the opportunities and risks at the intersection of human creativity and artificial intelligence. Converge-Con features free showcases, conversations and hands-on strategy sessions looking at how generative tools are reshaping everything from art to ethics. More information and registration at converge-con.ai.
By Xinyan (Eva) Zhao, Assistant Professor
Generative AI (GenAI) is no longer just a buzzword — it’s becoming a core tool for research innovation at UNC Hussman. Whether it’s transforming how we communicate strategically to inform and persuade audiences or how our students learn to navigate an evolving media landscape, GenAI is creating exciting new possibilities. As I have explored its potential in the past two years, I’ve found that GenAI is not just a tool; it offers new ways of thinking about both research methods and theories, helping us discover better approaches and expand the theoretical horizon. At UNC Hussman, researchers are using AI to make a real difference — not just in their research projects, but also in the classroom, where students are learning how to apply these cutting-edge tools to tackle real-world problems.
In my research, GenAI is enabling me and my colleagues to tackle challenges in disaster, crisis and health communication. We’ve developed and tested GPT-4-powered chatbots that help diverse ethnic communities prepare for hurricanes, and we’re integrating GenAI tools into our “MEJO 379: Advertising and Public Relations Research Methods” curriculum. While these innovations begin to reveal the potential of GenAI in research and teaching, we’re only scratching the surface of what it can achieve. As technologies continue to evolve, we remain committed to broadening their application to more diverse communication contexts and challenges. Our goal is not only to address socially significant issues but to empower students and communities by promoting inclusive and ethical AI use.
One of the most promising applications of GenAI in my research focuses on using GenAI-powered chatbots to improve disaster preparedness communication for diverse communities. AI agents powered by models like GPT-4 have the potential to transform how emergency information is delivered, particularly in regions like Florida, North Carolina and Texas. Traditional disaster communication methods often rely on one-sizefits-all messaging, which may not resonate with the varied needs and cultural backgrounds of local residents. This is where GenAI steps in.
Our project focuses on culturally tailoring hurricane preparedness information through prompted GPT-4 chatbots for Black, Hispanic and Caucasian residents. These GenAI chatbots incorporate cultural elements such as familiar names and phrases, multilingual support and cultural values to address community-specific concerns. For example, a chatbot tailored to Hispanic users can provide bilingual support, highlight Hispanic-specific disaster resources and emphasize cultural values like family and collectivism. For other groups, the chatbot can focus more on delivering general disaster information. These chatbots also varied in tone — some were informal and conversational, while others maintained a more formal, authoritative tone. This variation allowed us to experimentally test and identify the most effective communication strategies for different cultural groups.
We conducted an experiment with 441 participants from different racial and ethnic groups in Florida. In this study, we divided participants into separate groups, with each group interacting with a different version of chatbots that varied in tone or cultural tailoring. This approach enabled us to observe how changes in chatbot configuration influenced participants’ perceptions of the chatbots and their sense of preparedness to take action afterward. Since each group only experienced one version of the chatbot, we were able to compare the responses across the different groups. The results were significant. Chatbots that reflected the cultural background of users were seen as more credible. This, in turn, led participants to feel more confident in sharing and acting on the disaster preparedness information. Additionally, while chatbots with a more informal tone were viewed as friendlier, chatbots with a formal tone were perceived as more credible, increasing participants’ trust and encouraging them to act on the preparedness information. This points to the need to explore more nuanced strategies beyond tone to humanize chatbots for specific groups in specific contexts.
This research is among the first to explore the integration of GenAI with cultural tailoring in strategic communication and public relations. In doing so, we are pushing the boundaries of how AI can be applied in critical, high-stakes contexts like disasters and crises. The insights gained show AI’s potential not only to make communication more interactive and engaging, but also to foster a more inclusive approach that addresses the needs of vulnerable, diverse populations. By personalizing communication and making it more human-like, these chatbots could bridge the gap between authorities and communities, empowering individuals to take more informed and timely actions during disasters.
In response to the growing demand for AI proficiency in the field of advertising and public relations, our “MEJO 379: Advertising and Public Relations Research Methods” course now integrates GenAI tools into its curriculum, supported by UNC’s CFE/ Lenovo Instructional Innovation Grants. This initiative started in fall 2024 and introduces students to critical AI applications, ensuring they not only understand how to use AI in different ways but also how to assess its strengths, limitations and ethical implications.
One key component involved comparing traditional and AI-based qualitative data analysis.
“The insights gained show AI’s potential not only to make communication more interactive and engaging, but also to foster a more inclusive approach that addresses the needs of vulnerable, diverse populations.”
Students worked in two groups to analyze their focus group data: one using traditional thematic analysis and the other employing a customized GPT-4 or Microsoft Copilot. Through this process, students gained valuable insights into how AI influences research processes, learning how AI can augment human analysis while also recognizing the biases or gaps in AI-generated data interpretation.
The next component focuses on using GenAI for communication strategy development. Students learn AI prompting techniques to generate and refine communication strategies tailored to brand objectives and target audiences, followed by designing survey questions to test these strategies. In the final component, students will engage in AI-driven image generation for A/B testing, using tools like Adobe Firefly and Playground to create and compare visual content, and learning advanced techniques like prompt-to-prompt editing.
Our MEJO 379 class teaches students the technical skills necessary to work with GenAI but also challenges them to think critically about how AI influences strategy, content, analysis and decision-making. Through immersive, hands-on experience, students are prepared to lead the charge in the rapidly evolving field of strategic communication. With practical AI knowledge and a deep understanding of its implications, they become adaptable, innovative professionals, ready to tackle the complex challenges of tomorrow’s communication landscape.
