My teammate is an AI : evaluating generative AI in game asset creation through a post-gamejam study

Abstract

This study explores the integration of generative AI into game development, focusing on its use during a university-led game jam involving undergraduate students. The research investigates whether generative AI holds significance for the future of game creation - highlighting its potential or otherwise to transform asset production, team dynamics, and creative workflows. It examines how AI-driven tools such as ChatGPT and DALL-E impacted the development process, enabling hybrid roles and streamlining production tasks that involve repetitive visual patterns or templated interface elements. However, limitations emerged, including asset reliability, quality, and creative customisation challenges. Finding what this means for the broader industry while emphasising the importance of refining AI tools to balance efficiency with artistic integrity. A significant number of participants (46%) reported notable learning gains, reflecting improved technical skills and a deeper understanding of AI's role in creative processes. These results point to broader questions for future policy, including intellectual property, ethical data use, and inclusive development - although such issues were not explicitly raised by participants during the study. This research offers insights for advancing AI-human collaboration models to support sustainable, innovative, and culturally diverse game production.