Published • loading... • Updated
Study: AI coding tools do not write accessible code
Researchers found 35 accessibility issues per page on average, showing AI-generated code often misses basic standards for users with disabilities.
- A recent study by AudioEye found that AI-generated websites contain numerous accessibility errors per page. These failures often prevent users from completing basic tasks like checking out or signing up.
- Development teams increasingly rely on AI to ship code faster, assuming tools handle accessibility standards. However, these models lack specific accessibility expertise, producing code that remains unusable for assistive technology users.
- Despite these risks, 81% of business leaders and developers express confidence in AI-generated code for production. Testing revealed that sites failed in identical ways, including inaccessible forms or keyboard navigation errors.
- Coding failures create significant legal risks for e-commerce companies, as inaccessible sites often trigger demand letters or lawsuits. When customers cannot navigate a store using assistive technology, they abandon the site, directly impacting revenue.
- Experts warn that shipping automated code without rigorous testing leaves companies vulnerable to user exclusion and legal challenges, even as AI coding tools accelerate development.
Insights by Ground AI
36 Articles
36 Articles
Coverage Details
Total News Sources36
Leaning Left4Leaning Right0Center32Last UpdatedBias Distribution89% Center
Bias Distribution
- 89% of the sources are Center
89% Center
11%
C 89%
Factuality
To view factuality data please Upgrade to Premium
