AI Coding Tools Fail to Deliver Accessible Websites, AudioEye Study…

Event summary

AudioEye's study found 306 distinct accessibility issues across 15 AI-generated websites, with 91% rated medium or high severity.
81% of developers trust AI-generated code for accessibility, despite 73% reporting increased accessibility issues post-adoption.
46% of surveyed teams received demand letters or lawsuits, with 71% attributing issues to AI-generated code.
AI tools averaged 55 accessibility issues per page, nearly matching the 62-issue average of non-AI sites in AudioEye's 2026 Digital Accessibility Index.

The big picture

AudioEye's findings highlight a critical gap in AI-generated code, where speed and efficiency come at the cost of accessibility compliance. As AI tools are increasingly adopted for web development, the disconnect between developer confidence and actual accessibility performance poses significant legal and operational risks. This study underscores the need for specialized solutions that combine AI automation with human expertise to ensure digital accessibility.

What we're watching

Legal Exposure: The pace at which organizations relying solely on AI coding tools accumulate legal exposure from accessibility lawsuits.
Industry Standards: Whether AI tools can evolve to meet WCAG 2.2 AA standards without human oversight or supplementary solutions.
Market Response: How competitors will address the accessibility gap in AI-generated code, and whether AudioEye can capitalize on this market need.