Towards accessible website design through artificial intelligence: A systematic literature review

Web accessibility ensures that individuals with disabilities can access, navigate, and interact with online content effectively. Despite the availability of the Web Content Accessibility Guidelines (WCAG), significant barriers persist, largely due to the complexity of their implementation. Artificial intelligence (AI), particularly machine learning models, has emerged as a promising avenue to address these challenges, offering solutions for evaluation, correction, and content generation. This study aims to systematically review the intersection of web accessibility and AI by evaluating how AI-based methods enhance compliance with accessibility standards over the period 2019–2025, assessing their efficacy and alignment with WCAG principles. A systematic literature review (SLR) was conducted in three phases: planning, execution, and reporting. Research questions were formulated guiding the selection of search terms and strategies. A systematic search process was implemented, complemented by a snowballing technique to ensure comprehensive coverage of relevant studies. The quality of selected studies was rigorously assessed using predefined criteria, and data extraction was carried out following established best practices. The analysis combined narrative synthesis for qualitative insights and bibliometric techniques for quantitative evaluation . From 277 studies, 31 were identified as relevant, highlighting AI’s primary contributions to generating alternative text for images, automating compliance assessments, providing correction suggestions, and designing alternative interfaces to enhance accessibility. Advances in large language models (LLMs) further demonstrate their potential for facilitating the creation of accessible content. AI presents significant potential to improve web accessibility by streamlining evaluation processes and supporting the creation of accessible content. However, further research is needed to address limitations such as inconsistent compliance and the lack of focus on non-visual disabilities. These findings underline the importance of leveraging AI to facilitate inclusive web design practices while ensuring adherence to accessibility standards. • A SLR reviewing of 277 studies; 31 key works on AI methods for web accessibility. • Guide for systematic, rigorous SLRs aimed at novice web accessibility researchers. • Mixed-method SLR for AI-driven advancements in web accessibility including tools. • AI helps generate accessible content, but gaps remain beyond specific code cases. • Available dataset from the SLR to support research on AI and web accessibility.