ICAT Playdate — Effect of Integrating GenAI on Collaborative Learning and Problem Solving

With Generative AI (GenAI) becoming increasingly prevalent in workplaces, concerns have emerged about its impact on human collaboration. Specifically, does the presence of AI chatbots in group problem-solving fundamentally change how humans collaborate? In professional settings, individuals may rely more on AI assistants for information gathering and idea generation, potentially reducing direct human collaboration. While prior research has examined human-AI collaboration at the individual and organizational levels, little attention has been given to how humans interact with AI within the same group when solving complex problems. Our study investigates how the use of AI chatbots affects collaborative learning—the extent to which individuals share information about good solutions—and solution exploration—the development of novel solutions that previously did not exist within the organization. To explore these dynamics, we conduct a series of experiments with three conditions: (1) a control group, where the AI chatbot functions as a passive information storage tool; (2) Treatment 1, where each individual uses their own AI chatbot while collaborating with another human participant; and (3) Treatment 2, where participants share a single AI chatbot and can view each other’s queries and AI-generated outputs. Through this study, we aim to contribute to the human-AI collaboration literature by deepening our understanding of AI's role in group decision-making, knowledge exchange, and problem-solving.