Educational fields are adapting to rising use of generative artificial intelligence (genAI). We aim to discover genAI impacts in the field of transformative futures innovation education and to offer insights for further research and practitioners in associated fields. We explore master’s level students’ genAI prompts-answer pairs from IdeaSquare Planet (I2P) programs using content analysis to study self-organized, interdisciplinary students working on “wicked” problems related to the Sustainable Development Goals (SDGs) using futures and systems-thinking tools. We find students use genAI mainly for suboptimal tasks, e.g., calculations and quantitative information retrieval. In addition, students demand unambiguous answers from genAI to uncertain problems, going against the I2P goal of teaching students to navigate wicked problems. The impacts of genAI in such educational contexts are empirically unstudied and this research paves way for future research by forming testable hypotheses based on the findings on topics such as students’ discomfort with complexity and uncertainty.