Generative AI for Narrative and Character Design in Video Games

Sammanfattning

This thesis examines how useful generative artificial intelligence (GenAI) is in producing narrative and character designs for video games that adhere to domain-specific requirements. The rapid growth in the quality, capacity and usage of GenAI has brought questions about its usefulness. The thesis is a qualitative, semi-systematic literature review. Nine peer-reviewed contemporary papers published 2024 or 2025 were analysed using thematic analysis. The goal of the thematic analysis was to provide an answer for the research question: “How useful is generative AI in producing character and narrative designs for video games that adhere to domain-specific requirements?”. In this context, domain-specific requirements refers to the quality of the generated storylines, characters and game assets. The review reveals that using GenAI allows for adaptive narratives, enhanced production efficiency, and supports creative processes. On the other hand, GenAI seems to struggle with story coherence, character depth, asset quality, and representational bias. This implies that while GenAI is a useful development tool, requirements are not met independently and require human involvement in the workflow in order to do so.