Do Accessibility Overlays Work? What the Evidence Says

The pitch is irresistible if you are a business owner who has just received a demand letter. Add one line of JavaScript to your site and an overlay widget will detect and fix your accessibility problems automatically, making you compliant without touching your code. Some vendors have offered legal support or indemnity alongside it.

The question of whether this works is unusually well settled — not by opinion, but by a regulator, by the accessibility profession, and by litigation data. Here is what each shows.

What an overlay actually is

Two things get bundled under the word "overlay", and separating them clarifies a lot.

The first is a user-facing toolbar: a floating accessibility icon that opens controls for larger text, higher contrast, a dyslexia-friendly font, a reading mask, and so on. This is a preferences panel.

The second is automated remediation: JavaScript that scans the page after it loads and attempts to fix accessibility problems — guessing at missing alt text, adding ARIA attributes, adjusting contrast, inferring form labels.

The first is largely harmless but mostly redundant. Browsers and operating systems already offer zoom, contrast controls and font settings, and they work across every site rather than one. Disabled users overwhelmingly use their own assistive technology, configured the way they like it, rather than learning a bespoke widget on each site they visit.

The second is where the claims are made and where the problems are.

What the FTC found

In January 2025 the US Federal Trade Commission announced action against accessiBe, one of the largest overlay vendors, with the final order approved in April 2025 requiring the company to pay $1 million, which may be used for consumer refunds.

The allegations came in two parts. First, on the product: the FTC alleged that accessWidget did not make all user websites WCAG-compliant, and that the company's claims to that effect were therefore false, misleading or unsubstantiated. Second, on marketing: that the company deceptively formatted third-party articles and reviews to appear as independent opinions while failing to disclose its material connections to the reviewers.

The order bars accessiBe from claiming its AI products can make websites WCAG-compliant unless it has evidence to support the claim, and requires disclosure of connections to endorsers.

This is a regulator formally finding that "one line of JavaScript makes you compliant" was not a supportable claim. It is the strongest single data point available on the question.

What the accessibility profession says

The Overlay Fact Sheet is a public statement opposing overlays as a compliance solution. It carries more than 1,000 signatories, and the composition matters more than the count: contributors and editors to the WCAG, ARIA and HTML specifications; contributors to the JAWS and NVDA screen readers; accessibility staff at Google, Microsoft, Apple, the BBC, Shopify, Target, CVS Health and eBay; disability rights lawyers; and disabled users themselves.

Its central claim is unambiguous: no overlay product on the market can cause a website to become fully compliant with any existing accessibility standard, and therefore none can eliminate legal risk.

The user data behind that position is equally direct. In WebAIM's survey of web accessibility practitioners, 67% rated overlays as not at all or not very effective — and among respondents who themselves have disabilities, 72% said not at all or not very effective, with only 2.4% rating them very effective.

What the lawsuit data shows

If overlays reduced legal risk, sites using them would be sued less. They are not.

In 2024, more than 1,000 businesses with an accessibility widget installed were sued anyway — over a quarter of all digital accessibility cases tracked that year. The pattern continued through 2025, with UsableNet reporting that widgets "did not slow litigation" and that an increasing number of suits targeted companies already using them.

That is not evidence that overlays cause lawsuits. It is evidence that they do not prevent them, which is the thing being sold.

Why the technology cannot deliver what is promised

This is worth understanding structurally rather than as a matter of vendor quality, because it explains why a better overlay would not solve it either.

Most of WCAG is not machine-decidable. Automated tools produce results for roughly a third of WCAG Level AA success criteria. An overlay is an automated tool running in the browser; it inherits that ceiling. It cannot know whether your alt text is meaningful, whether your reading order preserves meaning, whether a process can be completed, or whether an error message helps.

Guessing at meaning produces confident errors. An overlay generating alt text for an image cannot know why the image is on the page — whether it is decorative, whether the caption already covers it, whether it is a link. The result is plausible-sounding descriptions that are frequently wrong, which is worse than an empty attribute because it passes automated checks while still failing the user.

Repairs applied after load can conflict with assistive technology. Screen readers build their model from the accessibility tree. An overlay mutating that tree after the page has loaded can produce duplicate announcements, focus that jumps unexpectedly, and controls that announce one thing and do another. Some disabled users report actively worse experiences on overlay sites, and a number now use browser extensions specifically to block overlays.

The underlying code is unchanged. Turn the script off and every original problem is still there. That is the clearest indicator of what has been bought: a layer over the defect rather than a fix for it.

Where they might reasonably sit

A fair position rather than a purist one: a preferences toolbar is not fraud, and if you already have one, you do not urgently need to remove it. What you need to do is stop counting it as remediation.

The specific things to do are:

• Do not represent an overlay as making you compliant in your accessibility statement, marketing or procurement responses. That representation is exactly what the FTC acted on.
• Do not let it displace real work. The failures overlays claim to fix — alt text, labels, contrast, keyboard access — are the same failures a scan will show you and a developer can fix properly in the source, once, permanently.
• Test with the overlay disabled, because that is your actual site.

Overlay adoption is, for what it is worth, small: the HTTP Archive's analysis of millions of sites finds around 2% of desktop sites using one, and only about 0.2% of the top thousand.

What to do instead

The alternative is not more expensive than an overlay subscription in most cases, and it is durable.

Scan continuously to catch the machine-detectable failures — the same ones plaintiffs' firms screen for — and to catch regressions as your site changes. Fix them in the source, where the fix persists and cannot be turned off. Manually test your critical journeys with a keyboard and a screen reader, because that is where the other two-thirds of the standard lives. And publish an honest accessibility statement describing what you tested and what you know is still broken.

That approach makes no promise of immunity, because none is available. What it produces is a site that is genuinely more usable, a documented record of good-faith work, and no representation that a regulator could call unsubstantiated.

*The FTC order against accessiBe was announced 3 January 2025 and finalised 22 April 2025. Signatory count from overlayfactsheet.com, which listed 1,031 signatories when checked in August 2026. Practitioner figures are from WebAIM's Survey of Web Accessibility Practitioners #3. Litigation figures are from UsableNet's 2024 and 2025 digital accessibility lawsuit reports. Market share figures are from the HTTP Archive Web Almanac 2025.*