Volume of AI-Generated Code is Significantly Increasing "Accessibility Debt" Risk: Deque Systems

• 24 Jul 2026
• TM Team
• 0 Comments

A newly released survey and accompanying report from Deque Systems, the global digital accessibility leader advancing digital equality with innovative software solutions, highlights a significant paradox. With development teams using AI coding tools to release software faster than ever before, the number of digital accessibility failures is increasing.

This escalating rate, combined with intensifying regulatory pressure, is leaving many organizations exposed to significant legal, financial and reputational damage. Fortunately, there's still time to course-correct, even as enforcement and penalties for noncompliance are expected to pick up in the months ahead.

"We are at a genuine inflection point, where regulation, AI transformation and the state of digital accessibility are converging in ways that demand a new kind of response from engineering leadership," says Preety Kumar, CEO and Founder, Deque Systems. "It's time to seize the moment to improve web and mobile experiences for people with disabilities everywhere."

The survey found that 88 percent of software engineering leaders have high or very high trust in the accessibility of their AI-generated code. Despite this, 64 percent cite accessibility as a top driver of post-production rework. This aligns with a recent analysis finding that 95.9 percent of the top home pages worldwide contain detectable WCAG 2 failures, reversing a trend of improvement over the previous six years. This isn't altogether surprising. In the GAAD Foundation's AIMAC benchmark, 35 of 37 of the latest AI models (94.59 percent) produce multiple critical and serious issues by default when asked to generate HTML.

This situation is particularly pressing as we mark the first anniversary of the European Accessibility Act (EAA), which is expected to intensify regulations and penalties for noncompliance. Notably, U.S.-based companies that market and sell to EU consumers are also subject to these rules, underscoring the far-reaching implications for businesses worldwide.

Moreover, recent extensions by the Department of Justice on the ADA Title II deadlines do not mean that digital accessibility requirements will fade away anytime soon. The message from regulatory bodies is clear - organizations must act now to ensure their websites and mobile apps are accessible for all people, everywhere.

AI use has enabled development teams to dramatically accelerate release velocity, shipping substantially more code with ever-increasing speed. But digital accessibility is complex, and simply prompting AI to produce accessible code is not enough to ensure compliance and reduce risk. Compliance demands accuracy and proof, and releasing inaccessible products and services can lead to serious legal, financial, and reputational damage. Fortunately, with expert guidance and the right tools and practices, digital accessibility can keep pace.

Kumar encapsulates the industry's opportunity: "At this critical juncture, we must ensure that AI is a catalyst for advancing digital accessibility. As our recent reporting demonstrates, accessibility must be fully embedded into AI-assisted development workflows. When that happens, costs go down. Risk goes down. This is how we turn accessibility debt into accessibility dividend."