When:
Wednesday, September 30, 2026 - 14:05 to 14:25 CEST
Room:
Penn Room I&II
Tags:
development & coding, ai
Track:
SVG
user experience, accessibility & design

The Reality of Accessibility Automation

The Reality of Accessibility Automation

Daniel Angelov (danielangelov)

Lighthouse gives you 100. The European Accessibility Act expects more.

Prerequisite

Nothing, I'll explain everything from scratch.

Target Audience

QAs engineers, team leads, architects, business owners, compliance officers.

Outline

Accessibility automation is often seen as the silver bullet for building inclusive digital products. With tools like axe, Lighthouse, and CI-integrated scanners becoming standard in modern development workflows, many teams assume that automated testing ensures accessibility compliance. The reality, however, is more complex.
In this session, we'll explore the true capabilities and limitations of accessibility automation. We'll break down what automated tools can reliably detect, where they fall short, and why a significant portion of WCAG success criteria still require human judgment. Through practical examples, I'll highlight the critical gap between rule-based validation and real user experience - particularly for screen reader users, keyboard-only users, and people with cognitive disabilities.
The session also examines the emerging role of AI in accessibility testing. Can AI help bridge the semantic and contextual gaps that traditional automation cannot cover? Where does it add value, and where does it still fall short?
Attendees will leave with a realistic understanding for accessibility quality - combining automation, AI augmentation, and human expertise. This talk challenges common misconceptions and provides a practical framework for building scalable, responsible accessibility testing strategies in modern teams.

Learning Objectives

By the end of this session, participants will be able to:
Explain what automated accessibility tools can reliably detect - and what they fundamentally cannot
Describe why accessibility automation typically covers only 20-40% of WCAG success criteria
Distinguish between rule-based validation and real user experience validation
Evaluate where AI can meaningfully enhance accessibility testing and where human judgment remains essential

Experience level
Beginner