The Accessibility Widget You Installed Is Why You Got Sued
Somewhere right now, a founder is pasting a single line of JavaScript into their site's <head> tag, closing the ticket labeled "accessibility compliance," and moving on to the next sprint. The widget promises WCAG 2.1 AA compliance instantly — no audit, no remediation, no developer time. A little accessibility icon appears in the corner of the screen. It feels like the problem is solved.
It isn't solved. In a meaningful number of cases, that icon is the reason the lawsuit lands on their desk six months later.
Accessibility overlay widgets — AccessiBe, UserWay, AudioEye, and a handful of competitors — sell a version of accessibility that requires nothing from your codebase. You install a script, it scans your DOM at runtime, and it layers interface adjustments on top: font resizing, contrast toggles, an AI-generated pass at missing alt text, keyboard-navigation patches. The pitch is seductive because it's the opposite of what real accessibility work actually requires, which is going through your markup, your component library, and your interaction patterns and fixing what's broken at the source.
The lawsuit data says the opposite of what the widget promises
Here's the number that should have killed the overlay industry's marketing years ago: 22.6% of all digital accessibility lawsuits filed in the first half of 2025 named sites that were already using an accessibility overlay. Not sites with no accessibility effort at all — sites that had specifically paid for the product marketed as the fix. Accessibility.Works and legal-tracking firm data put the count higher across 2023–2024: more than 800 businesses using overlay widgets were sued despite having them installed, with 258 lawsuits specifically naming AccessiBe-equipped sites and 187 naming UserWay sites in 2024 alone.
Individual cases make the pattern concrete. Eyebobs settled a suit in October 2021 despite running AccessiBe. ADP's December 2021 settlement explicitly rejected AudioEye and AccessiBe as sufficient remediation — the settlement terms required actual code-level fixes, not overlay coverage. Quezada v. U.S. Wings and Angeles v. Grace Productions followed the same shape: overlay installed, lawsuit filed anyway, overlay found legally insufficient.
The regulatory response caught up in 2025. The FTC fined AccessiBe $1 million in January 2025, finalized that April, for false advertising and — this is the detail that matters — fake customer reviews inflating the product's effectiveness. That's not a company being sued for having imperfect accessibility. That's a company being fined by a federal agency for lying about what its product does.
WebAIM's data explains the mechanism, not just the outcome
The WebAIM Million report — an annual scan of the top one million website homepages — gives the technical reason the lawsuits keep landing. Sites running detectable overlay widgets average 36.9 to 50.5 accessibility errors per page, against an overall average of 56.8 errors across all sites scanned. That looks like an improvement until you read WebAIM's own methodology note: overlays that manipulate test results were disabled before testing. In other words, the modest error reduction reflects the underlying quality of sites that happen to install overlays, not anything the overlay itself repaired. Strip out the overlay's own attempts to game automated scanners, and the errors it was supposed to fix are still there.
The technical reason is straightforward once you know how screen readers actually work: they parse a page's HTML and ARIA structure on load, before most overlay scripts have finished executing their DOM adjustments. A screen reader user often hits the broken structure first and the overlay's patch second, if at all. The overlay is renovating a house after the inspector has already walked through and failed it.
This is why the disability community's rejection of overlays has been so consistent and so loud. The National Federation of the Blind — the largest blind advocacy organization in the US — banned AccessiBe from its 2021 national convention outright, calling the products harmful rather than merely insufficient. The Overlay Fact Sheet, an advocacy initiative founded by accessibility consultant Karl Groves, has collected more than 1,031 signatories, including disability advocates and engineers from Google, Microsoft, and Apple, all committing to the position that overlays "are not an effective means of ensuring accessibility." That's not a fringe objection. That's most of the field that actually does this work for a living, on record, saying the product category doesn't do what it claims.
So actually — the widget is a liability shield with a hole in it
The overlay pitch works because it reframes accessibility as a checkbox problem: install the thing, get the certificate, move on. But accessibility isn't a state a script can retrofit onto a page after the fact — it's a property of how the page was built, which means it has to be addressed in the markup, the component library, and the design tokens, not layered on top of them at runtime. Related design-system work on dark mode implementations too often shipping without the accessibility considerations baked in from the start runs into the same root cause: bolt-on accessibility measures behave differently from accessibility that was part of the build from the beginning, and the gap is exactly where lawsuits and screen-reader failures both live.
The uncomfortable part for teams that already installed an overlay: it's not neutral. It's not "better than nothing while we get to real remediation." The FTC fine and the lawsuit data suggest it can function as evidence that you knew accessibility was a legal exposure and chose the option that looked like compliance without doing the work compliance actually requires. That's a worse legal position than having done nothing, in some read of it — because doing nothing is a gap, and installing a widget that's been publicly and repeatedly shown not to work looks more like willful avoidance once it's in front of a judge.
If your accessibility strategy fits in one script tag, it isn't a strategy. It's a bet that no one representing a screen reader user will ever actually try to use your site — and 800 companies already lost that bet.