Abstract Artificial intelligence (AI) has enabled the rise of AI-powered avatar assistants–digital entities that facilitate interaction through natural language processing, speech recognition and real-time animation. These avatars are increasingly used in education, healthcare, customer service, entertainment and cultural heritage. This systematic literature review (SLR), following the PRISMA framework, analyses 35 studies published from 2020 to 2025 to classify avatar functionalities and identify application domains while exploring the current tools and challenges associated with them. Although the field shows significant growth and innovation, persistent challenges remain. By systematically organizing the current state of AI avatar research and development, this review outlines existing challenges, research gaps, discusses future directions and highlights a foundation for more inclusive, secure and effective deployment of AI-powered avatars. The insights presented aim to guide researchers and developers in understanding the current landscape and advancing the responsible use of AI-powered avatar assistants in immersive spaces.
Halim et al. (Thu,) studied this question.