Every few months an Australian business owner asks us the same question: "Can we just put one of those accessibility widgets on the site? The kind with the little wheelchair icon in the corner?" The honest answer is yes you can install one. It will not make your site compliant. It may make it worse. And as of 2025, telling a court you bought one is no longer a defence — the Federal Trade Commission fined the largest vendor a million US dollars for misrepresenting exactly that claim.

The opinion up front, because this matters: accessibility overlay widgets are a worse choice than doing nothing, in two specific ways. They give you false confidence, so you stop investing in the real work. And they actively break the experience for some screen reader users, who then have grounds for a discrimination complaint that wouldn't have existed if you'd simply left the site alone.

The nuance we'll get to: there is a narrow case where an overlay can be a temporary bridge during a planned remediation. We'll cover when that applies. It is not what the vendors are selling, and it is not what most Australian businesses buy them for.

What an accessibility overlay actually is

An overlay is a piece of third-party JavaScript that loads on your site and tries to "fix" accessibility issues at runtime. The major vendors are accessiBe, UserWay, AudioEye, and a long tail of smaller copycats. They typically install with one line of code and add a floating widget — usually a wheelchair or accessibility icon — that lets users toggle things like high contrast, larger text, dyslexia-friendly fonts, or keyboard navigation modes.

Behind the scenes, the overlay scans your DOM and applies fixes it thinks will help. Missing alt text gets AI-generated descriptions. Forms without labels get inferred labels. Insufficient colour contrast gets overridden. Buttons that aren't keyboard-focusable get patched.

The marketing pitch is irresistible to a business owner who's just been told their site needs to comply with the Disability Discrimination Act. One line of code. WCAG AA compliance. AUD $50 a month. The vendor handles everything. Why would anyone build accessibility into the site itself when you can just bolt this on?

The pitch is a fiction. Here's why.

The technical problem

Accessibility is, fundamentally, about the underlying code being correct. A screen reader reads the document object model. A keyboard user navigates by tab order set by the DOM. A voice control user activates elements by their accessible name in the DOM. The overlay can't rewrite the DOM in a way that screen readers reliably understand.

Take alt text. If your image has no alt attribute and the overlay's AI generates "person standing in front of building," that string may or may not be announced by a screen reader, depending on how the overlay injects it. Some screen readers honour aria-label overrides. Some don't. Some announce both the original empty alt and the injected one. The user experience is unpredictable in a way that the vendor's demo doesn't show.

Take form labels. If your form input has no associated label, the overlay tries to infer one from nearby text. Sometimes it gets it right. Sometimes the inferred label is wrong, and now a blind user is filling in the "First Name" field with their phone number because the overlay misread the layout. There's no way for the user to know the inferred label is wrong.

Take keyboard navigation. If your custom dropdown is built with non-focusable divs, the overlay adds tabindex="0" so it can be reached. But it doesn't add the keyboard event handlers, the ARIA expanded state, or the focus management on close. The element is now reachable but still broken. The user has been led further into a trap rather than away from it.

This is why the Web Accessibility Initiative, the National Federation of the Blind, and every credible accessibility consultant in the industry has signed the Overlay Fact Sheet saying overlays don't work.

The legal problem in the United States

There were 4,000 ADA Title III lawsuits filed against US businesses for inaccessible websites in 2024. UsableNet's annual report found that 25% of those lawsuits targeted sites with an accessibility overlay installed. That share grew to nearly a third in 2025. Plaintiffs' lawyers specifically search for sites with overlay code as easy targets — the overlay's presence doesn't help the defendant and the existence of accessibility barriers despite the widget is easier to demonstrate.

In April 2025, the Federal Trade Commission fined accessiBe USD $1 million in a settlement that prohibited the company from claiming its product makes any website WCAG compliant. The FTC said accessiBe "misrepresented the ability of its AI-powered web accessibility tool to make any website compliant with the Web Content Accessibility Guidelines for people with disabilities" and also penalised them for fake reviews. This is now part of the public record and admissible in any future case.

UserWay was sued in a 2024 class action filed in Delaware after their widget failed to prevent a flower delivery company being sued by a disabled customer. The class action alleges UserWay promoted its product as foolproof compliance and failed to deliver. The case is ongoing.

What this means under Australian law

Australia doesn't have the same volume of accessibility litigation as the United States. The Disability Discrimination Act 1992 funnels complaints through the Australian Human Rights Commission's conciliation process before any court action. That means we don't see annual lawsuit counts, and most matters resolve confidentially.

But the legal framework treats the overlay claim identically. The DDA test is whether your service is actually accessible to a person with a disability. Not whether you bought a product that claims to make it accessible. Not whether you have a wheelchair icon in your footer. The test is whether a blind user can actually complete a purchase, a deaf user can actually understand a video, a motor-impaired user can actually navigate a menu.

If the Commission's conciliator asks "did your customer successfully order from your site?" and the answer is no, the overlay you installed doesn't help. It's evidence you knew about your obligation and chose a path that didn't work.

The AHRC's April 2025 guidelines on equal access to digital goods and services — the same guidelines we covered in our DDA explainer — go further. They warn explicitly about AI tools and automated accessibility products as a source of risk if they aren't designed inclusively. The regulator has read the FTC ruling. So has every accessibility lawyer in the country. The overlay defence is dead in Australia in 2026.

Why some users actively hate them

There's a second-order problem nobody mentions in the sales pitch. Accessibility overlays interfere with the assistive technology that screen reader users have already configured.

Most blind users in Australia use JAWS, NVDA, or VoiceOver, with personal settings tuned over years. Speech rate, voice, verbosity, navigation shortcuts — all configured to their preference. When they land on your site and the overlay loads, it often tries to do things their screen reader already does, in a different way, with different keyboard shortcuts, conflicting with their setup.

The WebAIM survey of screen reader users in 2023 found that 72% of respondents had a negative view of overlay widgets and 67% said they made sites harder to use. Those are users with lived experience. The overlay vendors' own user-study claims do not survive contact with this data.

It's worth understanding the emotional component too. A blind user who has spent years configuring their screen reader experiences your overlay as your site shouting at them: "Don't worry, we've got accessibility handled for you!" — when they had it handled before you intervened. The widget is condescending to the people it's marketed as helping. That hits differently when you understand it.

The narrow legitimate case

There is exactly one situation where an overlay can be defensible: as a temporary bridge during a planned, budgeted, scheduled remediation project.

Concretely: you've identified that your existing site has serious accessibility issues. You've engaged a developer to fix the underlying code. The fix will take 8 weeks. You install a free overlay for the 8-week window so that some users get some benefit while the real work proceeds. At the end of week 8, you remove the overlay because the fix is done.

That's it. That's the case. Any other use — as a permanent solution, as a substitute for real work, as a compliance shield — is the use the FTC penalised. And even in the bridge case, the overlay is supplementary, not a substitute. You still publish an accessibility statement disclosing the known issues and a contact path. You still tell users with disabilities they can phone or email if the site doesn't work for them.

We've never recommended this case to a client in practice, because the budgets that go to overlays would always be better spent on the underlying remediation. But it's the only honest place an overlay belongs.

What the money would actually buy

A typical accessibility overlay costs AUD $50–$100 per month per site. That's $600–$1,200 a year. Over three years, $1,800–$3,600.

For the same money, an Australian SMB can:

Have a manual WCAG 2.2 AA audit done by a specialist, generating a prioritised remediation list (~$2,000)
Pay a developer two days to fix the most critical issues (~$2,000)
Publish a real accessibility statement
Train staff who update the site
Schedule a follow-up audit in 12 months

That's actual compliance work, not theatre. It compounds — the next time you audit, you're starting from a better baseline. The overlay subscription, by contrast, compounds nothing. You're renting a sticker.

What the overlay vendors are likely to argue back

For completeness, the strongest version of the pro-overlay argument we've seen, and our honest read on each.

"Overlays are improving and AI is getting better." Partially true. The AI behind alt text generation and label inference is better than it was three years ago. It is still nowhere near reliable enough to pass an actual screen reader test, and the gap between "the demo video looks good" and "a real user can buy from your store" remains large. AI improvements have shrunk the gap, not closed it.

"It's better than nothing for businesses that can't afford a full rebuild." This is the most sympathetic version of the argument and it's still wrong. The cost of a remediation pass on an existing site is in the same order of magnitude as three years of overlay subscription fees. The overlay buys you the appearance of action without the substance. For small businesses without the budget for a full audit, a free WAVE scan and an internal effort to fix the top issues delivers more accessibility improvement than any overlay.

"Some users find the toolbar useful." A handful do. Most don't. The WebAIM survey data is clear that the overall user experience for screen reader users worsens when overlays are present. Even if a small subset benefits from a contrast toggle, your site can offer those toggles directly without the overlay's other behaviour.

The conversation to have with your current vendor

If you're locked into an overlay contract, the relevant question to your vendor is direct. Ask them to show you, in writing, a list of specific WCAG 2.2 AA success criteria their product addresses, and which ones it doesn't. Ask whether their product would have prevented the lawsuits filed against the 1,416 companies using overlays that were sued in 2025. Ask what they cover if your business is the next one to receive an AHRC complaint.

The honest answers will be uncomfortable for the vendor. The unhonest answers — "we're WCAG compliant" or "we offer full ADA compliance" — are precisely the language the FTC penalised accessiBe for using.

Where this leaves you

If you currently have an overlay installed, the right move depends on your situation. If your site is otherwise accessible and the overlay is just sitting there doing nothing, removing it is low-risk. If your site has real issues and the overlay is the only thing pretending to address them, removing it is honest but exposes you — you need a remediation plan in motion before you do it.

The cost of doing it properly is real but bounded. The cost of doing it via overlay is also real and the protection it offers is now publicly worthless. The arithmetic only points one way.

If you'd like us to audit what's actually accessible on your site, build a remediation plan with real numbers, and give you a path off the overlay if you have one, book a free audit. We'll be straight with you about what's salvageable, what needs rebuilding, and what it actually costs.

Why Accessibility Overlays Fail: The $1M FTC Truth