Error-Prone AI Accessibility Tools Reportedly Lead to Navigation Issues for Blind Internet Users

April 7, 2024

AI-powered accessibility overlays on websites have been reportedly mislabeling or misinterpreting content, causing navigation difficulties for visually impaired users and others with disabilities. These tools allegedly interfere with screen readers and provide inaccurate descriptions, thereby complicating navigation. The inconsistency observed in these tools has triggered legal action as the companies responsible strive for compliance with accessibility regulations.

Contribute to Project Cerebellum and help establish guardrails for AI that prioritize trustworthy and accessible experiences. Join us today at JOIN US.

This incident highlights the importance of the HISPI Project Cerebellum TAIM's Measure function, emphasizing the necessity of evaluating and improving the accuracy of AI accessibility tools.

Matched TAIM controls

Suggested mapping from embedding similarity (not a formal assessment). Browse all TAIM controls

Alleged deployer
zara, pemex, lvmh, capita, companies-using-ai-based-accessibility-tools
Alleged developer
equalweb, userway, developers-of-ai-based-accessibility-tools
Alleged harmed parties
blind-people, visually-impaired-people, jakob-rosin

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/830

Data source

Incident data is from the AI Incident Database (AIID).

When citing the database as a whole, please use:

McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.

Pre-print on arXiv · Database snapshots & citation guide

We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.