Generative artificial intelligence (AI) is changing game development by automating the creation of two- and three-dimensional assets, adaptive worlds, and personalized stories. This narrative review analyses peer-reviewed work from 2018 to 2024 that integrates generative models—GANs, VAEs, large language models, and diffusion networks—into the game-design pipeline. The evidence shows that these tools cut production time and cost while allowing independent studios to achieve asset quality once limited to triple-A developers. Paired with virtual or augmented reality, generative AI supports play that reacts to players in real time and deepens emotional engagement. Key challenges remain: maintaining game balance, avoiding uniform content, reducing algorithmic bias, and clarifying intellectual-property rights. The review therefore proposes hybrid human–AI workflows, routine bias audits, and domain-specific metrics for technical accuracy and narrative consistency. Embedding clear ethical standards and broad-access strategies can help generative AI elevate video games as platforms for cultural expression, social connection, and experiential learning.
Alghamdi et al. (2025) studied this question.