In our recent Marketing Research Seminar, we had the pleasure of hosting Professor Laura Brandimarte (University of Arizona), who held a thought-provoking presentation on the evolving challenges and opportunities posed by Generative Artificial Intelligence (GenAI). Together with her co-authors, Prof. Brandimarte explored GenAI’s growing ability to produce text that closely mimics human writing, raising critical concerns around the misuse of such tools, particularly in the realm of academic and content plagiarism.

Drawing inspiration from the creativity literature on divergent thinking, a novel metric called DAT(Sent), which measures semantic dissimilarity between sentences in a document, was introduced. The core insight of the study is that human-written texts tend to show greater sentence-level semantic variation compared to AI-generated content. As a result, a new detection tool was developed in order to assess both sentence- and word-level dissimilarities to identify AI-generated text; it demonstrates reduced bias in identifying AI-written content, marking a notable step forward in responsible AI governance.

Apart from that, Prof. Brandimarte discussed how human creativity can be enhanced by GenAI, and what role the overall AI “acceptance” plays in the creation of texts and the development of advertising materials.

We express our gratitude to Prof. Brandimarte for the insightful presentation that deepens the understanding of AI’s creative capacities and underscores the importance of developing robust safeguards as these technologies continue to advance.