Logo image
Evaluating AI-Powered Website Accessibility Overlays
Conference proceeding

Evaluating AI-Powered Website Accessibility Overlays

Parker Hartman and Tim Gorichanaz
Proceedings of the 27th International ACM SIGACCESS Conference on Computers and Accessibility, pp 1-4
26 Oct 2025

Abstract

Human-centered computing -- Accessibility design and evaluation methods Human-centered computing -- Accessibility systems and tools
This paper evaluates the effectiveness of current AI-powered website accessibility overlays. In this study, two HTML-based websites with different layouts were built to include a range of known accessibility errors. The sites were then equipped with three different overlays, which were tested through both automated and manual means. The results show that while automated testing reported improved accessibility with the overlays, manual testing (via keyboard and screen reader) revealed that major accessibility errors remain. The accessibility overlays show the most success with alt text provision and form field corrections, albeit with substantial drawbacks. Contrary to the claims of their providers, automated overlays alone cannot mitigate legal risk for inaccessibility; manual human testing and solution development is still needed to address most website accessibility errors.

Metrics

Details

UN Sustainable Development Goals (SDGs)

This publication has contributed to the advancement of the following goals:

#3 Good Health and Well-Being

InCites Highlights

Data related to this publication, from InCites Benchmarking & Analytics tool:

Web of Science research areas
Computer Science, Artificial Intelligence
Computer Science, Interdisciplinary Applications
Computer Science, Theory & Methods
Telecommunications
Logo image