Pa. — Practically or theoretically, whether teaching a classroom full of Penn State students or pondering concepts with a smaller group of international colleagues, Heather Shoenberger brings curiosity and optimism as she considers the challenges and opportunities related to artificial intelligence (AI)
As an associate professor of advertising in the Donald P. Bellisario College of Communications at Penn State, she’s providing information, shaping perspectives and developing approaches — all while helping wider audiences better understand AI and how it could shape the future.
Classroom creation
Shoenberger co-created, with Mehrdad Mahdavi, an associate professor of computer science and engineering at Penn State, a three-credit undergraduate course. She has taught COMM 200N Generative AI for All since the fall semester of 2024.
It’s cross listed in the Bellisario College and the College of Engineering and reaches its enrollment cap each semester with ease. Sections are taught in both colleges.
“When the technology started to emerge, it was clear it was going to affect a lot of what we do here. We had no idea exactly what would happen or how it would manifest, but it was going to be a significant change,” Shoenberger said. “To really serve our students, we have to be able to tell the story of a technology that could cause so many disruptions, good and bad. We want our students to be able to understand how generative AI works, what ethical issues it creates, sociopolitical issues it invokes and so much more.”
Shoenberger said she works hard to meet her students’ expectation, and she’s able to challenge and inform them because of how she gets ready for class — often in real time.
“It’s a significant amount of work, and you always prepare to be ahead,” she said. “The goal is to convey high-level information, help them understand how the technology works and how ethical and societal implications overlap. At the same time, the technology is always changing and there are never-ending opportunities to keep the content fresh.”
Class always leads to important conversations, Shoenberger explained.
“I appreciate the collaborative discussion, and I always enjoy a good conversation,” she said.
Relational research
Those same traits led to one of Shoenberger’s most recent research publications, an article that introduced the human-AI asymmetric relationship theory (HAART). The theory outlines a structural mismatch in how humans and conversational AI consider their exchanges. While AI conversational agents can display relational characteristics, they cannot experience relational benefits or commitment. While users may perceive a mutual aspect to the relationship, none exists.
The paper was published in a special AI-focused issue of the <a href="https://doi.org/10.1080/00913367.2026.2735425″ rel=”nofollow noopener” target=”_blank”>Journal of Advertising, the most prestigious journal in Shoenberger’s field of study. After the journal’s editor asked her to contribute to the special edition, Shoenberger worked with four colleagues from around the world — Andrés Gvirtz of King’s College London, Sean Sands of CQUniversity in Australia, Veronica Thomas of Old Dominion University and Lotte Willemsen of the University of Amsterdam.
Shoenberger had long been curious about investigating the relationship between people and conversational AI, especially the perception of reciprocity as familiarity and technology evolved. The idea resonated with the group, Shoenberger said, particularly the potential of examining the topic from a consumer relationship perspective.
“I so badly wanted to pursue this idea of how AI mimics interpersonal communication, and the team just clicked right away,” Shoenberger said. “It was the most fun project I’ve ever worked on.”
HAART provides a framework for exploring and understanding human-AI relationships, especially in the context of advertising. According to Shoenbeger, the theory posits that AI is unable to experience relationships but can simulate emotional responsiveness, intentional commitment, attachment and intimacy. These simulated behaviors can mask their incapacity for relational reciprocity, leaving consumers under the perception that AI can develop a relationship — which she said could leave consumers vulnerable to dependence, risk and control.
The theory might ultimately lead to more transparency about such interactions and, potentially, regulation, Shoenberger said.
“I think we all were very hopeful that this would be helpful for future research in the field of advertising,” she said.
Last Updated November 10, 2026
