PURPOSE: With the rapid advancement of artificial intelligence, applications in accessibility design for persons with disabilities are continually expanding. However, the related research remains fragmented, with insufficient evidence regarding the practical implementation of personalised, intelligent accessibility design and limited examination of emerging trends and challenges associated with design efficiency, user experience optimisation, and interdisciplinary integration. METHODS: This study utilises bibliometric analysis, drawing on literature from the Web of Science Core Collection database, and employs software tools such as Microsoft Excel 2021, VOSviewer, CiteSpace and Charticulator to examine the current state, research hotspots, and future development trends in this field. RESULTS: The results indicate a significant increase in the number of publications in this field between 1 January 2001 and 31 December 2025, particularly in terms of research collaboration between institutions in the United States and China. The findings further demonstrate that, in recent years, scholarly research on accessibility design for persons with disabilities has primarily focused on key themes such as "artificial intelligence," "assistive technology," and "deep learning." CONCLUSION: We constructed a global theoretical knowledge map and used it to synthesise fragmented evidence on AI-enabled accessibility design, revealing a paradigm shift from "single-sensory compensation" to "multimodal cognitive enhancement." Drawing on identified hotspots and structural gaps-particularly limited attention to multiple disabilities and insufficient social-ethical engagement-we propose targeted directions to better align technological development with the diverse needs of persons with disabilities.
Tao et al. (Wed,) studied this question.