Zoltan Majdik

My research examines the persuasive capacities of AI systems and language models, with a focus on how linguistic and rhetorical features of human <-> AI interaction shape model behavior and outputs. I use experimental methods to study both the computational mechanisms by which language models produce contextually responsive outputs and the behavioral dynamics of how people interpret and act on AI-generated content. My current work bridges computational and behavioral approaches to investigate how prompt-level linguistic variation relates to internal model processes, connecting discourse-level analysis with emerging methods in AI interpretability and alignment.

I'm a recent co-editor of the Rhetoric Society Quarterly special issue on Rhetoric & AI, and the co-director of NDSU’s Center for the Study of Digital Society.