This paper addresses the lack of dedicated Image Quality Assessment (IQA) criteria for AI-generated omnidirectional images, which are crucial for Virtual Reality (VR) and Augmented Reality (AR) applications. The authors establish a large-scale IQA database named AIGCOIQA2024, containing 300 omnidirectional images generated using five different AI models and 25 text prompts. They conduct a subjective IQA experiment to assess human visual preferences from three perspectives: quality, comfortability, and correspondence. The database is then used to evaluate the performance of state-of-the-art IQA models. The study highlights the unique distortions in AI-generated omnidirectional images and the need for comprehensive evaluation methods. The results show that current models struggle to handle this new task, particularly in balancing quality, comfortability, and correspondence simultaneously. The paper concludes by suggesting future research directions to improve the assessment of AI-generated omnidirectional images.This paper addresses the lack of dedicated Image Quality Assessment (IQA) criteria for AI-generated omnidirectional images, which are crucial for Virtual Reality (VR) and Augmented Reality (AR) applications. The authors establish a large-scale IQA database named AIGCOIQA2024, containing 300 omnidirectional images generated using five different AI models and 25 text prompts. They conduct a subjective IQA experiment to assess human visual preferences from three perspectives: quality, comfortability, and correspondence. The database is then used to evaluate the performance of state-of-the-art IQA models. The study highlights the unique distortions in AI-generated omnidirectional images and the need for comprehensive evaluation methods. The results show that current models struggle to handle this new task, particularly in balancing quality, comfortability, and correspondence simultaneously. The paper concludes by suggesting future research directions to improve the assessment of AI-generated omnidirectional images.