When Generative Artificial Intelligence meets Extended Reality: A Systematic Review

TL;DR

The article conducts a PRISMA-based systematic review of generative AI applications in Extended Reality (XR) from 2023–2025, extracting 26 relevant studies to map current use cases, models, and technical challenges. It reveals a dominance of VR/AR paradigms, with diffusion models and large language models (LLMs) playing prominent roles, and natural language input as the prevailing interaction modality. Key findings highlight design and education as the most active domains, frequent use of image and 3D content generation, as well as significant challenges in multi-modal integration, latency, system interoperability, AR pipelines, and safety/security concerns. The authors propose future research directions that emphasize deeper multi-modal integration, edge-aware and scalable architectures, standardized interfaces, and responsible AI governance to unlock more robust, immersive, and trustworthy generative XR experiences with real-world impact.

Abstract

With the continuous advancement of technology, the application of generative artificial intelligence (AI) in various fields is gradually demonstrating great potential, particularly when combined with Extended Reality (XR), creating unprecedented possibilities. This survey article systematically reviews the applications of generative AI in XR, covering as much relevant literature as possible from 2023 to 2025. The application areas of generative AI in XR and its key technology implementations are summarised through PRISMA screening and analysis of the final 26 articles. The survey highlights existing articles from the last three years related to how XR utilises generative AI, providing insights into current trends and research gaps. We also explore potential opportunities for future research to further empower XR through generative AI, providing guidance and information for future generative XR research.

Figures (8)

• Figure 1: PRISMA flow diagram for systematic review.
• Figure 2: Selected articles per year (N=26). We observe a huge surge in 2024. The collected article in 2025 is less than in 2024 because our search period stopped at Q1 2025.
• Figure 3: Frequently-used keywords in the included articles.
• Figure 4: Application domains in the included articles.
• Figure 5: Generating 3D models by MS2Mesh-XR: It is important to note that users with simple gestures and voice commands can generate 3D models in immersive environments.