Engineering8 min read
ADA website lawsuits: who gets sued, and for what
Over 2,000 filings in six months, mostly against small e-commerce, and Shopify is the most named platform. What the law requires, and why a widget is no defence.
If you sell online in the United States, the accessibility of your storefront stopped being a design question some time ago. It is now a litigation question, and the filing data says the defendants are mostly companies that would not describe themselves as large.
This is an engineering piece, not legal advice. What follows is what the public filing data shows, what the law does and does not currently require, and what work actually changes your position.
The numbers, and who is counting them
Counts differ a lot depending on methodology, so it is worth naming the source each time.
| Source | What it counts | Figure |
|---|---|---|
| EcomBack | Website-only cases, Jan to Jun 2026 | 2,045, up 1.54% year on year |
| Seyfarth Shaw | All federal ADA Title III filings, Jan to Jun 2026 | 5,006, up 9% |
| Seyfarth Shaw | All federal ADA Title III filings, full-year 2025 | 3,117, up 27% on 2,452 in 2024 |
| UsableNet | Projection for full-year 2026 | about 6,176 across federal and state courts |
The narrowest count, website cases only, still works out at more than eleven new filings every working day, holding at the record set in 2025.
Three things in that data matter more than the totals.
It is e-commerce. UsableNet puts 79% of 2026 digital accessibility lawsuits against e-commerce businesses.
It is not the Fortune 500. Around 68% of companies sued reported under $50 million in annual revenue. The plaintiff economics work precisely because the defendant is small enough to settle rather than litigate.
It is concentrated. Ten plaintiff law firms filed 80.59% of first-half 2026 cases, and thirty plaintiffs filed more than half of them. Four states, Illinois with 463, California with 424, New York with 399 and Florida with 333, accounted for 79% of filings. Illinois is now first, which is a change from the years when this was mostly a New York and Florida story.
That last point is the useful one. This is not a random risk spread evenly across the web. It is a small number of firms running a repeatable process against sites they can screen automatically.
Shopify is the most named platform
In the first quarter of 2026, by EcomBack's count:
| Platform | Lawsuits | Share |
|---|---|---|
| Shopify | 459 | 44.26% |
| Custom-coded | 324 | 31.24% |
| WordPress | 150 | 14.46% |
Read that carefully, because the obvious conclusion is the wrong one. Shopify is not less accessible than the alternatives. Shopify has an enormous share of exactly the kind of store these firms target: consumer retail, US-facing, under $50 million, running a purchased theme that nobody has audited.
A theme from the Shopify theme store is a starting point, not a compliance position. The parts that generate complaints are almost always the parts a merchant added or customised: the announcement bar, the cookie banner, the quick-view modal, the variant picker, the apps bolted into the cart.
What the law actually requires, which is less settled than you are told
Most writing on this topic states flatly that your website must meet WCAG 2.1 AA. That is not accurate for a private business in the US, and the difference matters when you are deciding what to spend.
Title II, state and local government. There is a binding technical standard here. The Department of Justice's April 2024 rule set WCAG 2.1 Level AA for web content and mobile apps, at 28 CFR 35.200(b). It explicitly covers third-party content delivered "through contractual, licensing, or other arrangements", so a city's payment app counts even when a vendor runs it. DOJ has since pushed the deadlines back by a year: large entities to 26 April 2027, smaller entities and special districts to 26 April 2028.
Title III, private businesses, which is you. There is no technical standard. DOJ has never issued one. It has said repeatedly that public accommodations must make their websites accessible, and courts, consent decrees and DOJ settlements have used WCAG as the benchmark, sometimes 2.0 AA and sometimes 2.1 AA.
So the honest position is this: no regulation names a version you must hit, and WCAG 2.1 AA is nonetheless the standard you will be measured against, because it is the one the complaints cite and the settlements adopt. You do not get to argue that the absence of a rule is a defence. You also should not let anyone sell you certainty that does not exist.
If you also sell into the European Union, the picture there is different and firmer: the European Accessibility Act has applied since June 2025, it binds non-EU sellers reaching EU consumers, and it does name a standard. We went through that in the European Accessibility Act and your Shopify store.
The widget is not a defence, and there is now data
The overlay pitch is that one line of JavaScript makes a site compliant. Two separate facts have settled this.
The regulator. In January 2025 the Federal Trade Commission brought a case against accessiBe over its accessWidget product. The complaint alleged the company marketed the tool as something that would automatically comply with WCAG 2.1 AA, that it failed to make basic components such as menus, headings, tables, images and recordings compliant, and that it introduced accessibility barriers on sites where it was installed. accessiBe settled for $1 million, without admitting the allegations, under a final consent order approved on 24 April 2025. The order permanently bars it from claiming its products make a website WCAG compliant, and from unsubstantiated compliance or legal protection claims.
Be precise about what that is. The FTC regulates advertising. The order says the vendor may not promise compliance. It is not a court holding that overlays violate the ADA.
The filings. This is the number that settles it in practice. In the first half of 2026, 496 ADA website lawsuits, 24.25% of the national total, were filed against websites that already had an accessibility widget installed. A quarter of all defendants had bought the product that was supposed to prevent this.
An overlay runs after the page loads, guesses at the semantics the markup failed to provide, and cannot repair a keyboard trap, a custom control with no accessible name, or a checkout step that only works with a mouse. Those are the findings that appear in complaints.
What actually changes your position
The work is unglamorous and it is in the markup, not in a script tag.
- Keyboard first. Every path that leads to money, which means search, variant selection, add to cart, cart drawer, checkout, must complete with a keyboard alone, in a visible focus order, with no trap. Test it by unplugging the mouse, not by reading a report.
- Accessible names on everything interactive. Icon-only buttons, the cart toggle, the quantity stepper, the close control on every modal. An unnamed button is the single most common finding and among the cheapest to fix.
- Modals and drawers done properly. Focus moves in, focus is trapped while open, Escape closes, focus returns to the control that opened it. Quick-view and cart drawers are where themes fail most often.
- Forms with real labels and real errors. A placeholder is not a label. Errors must be announced, associated with the field, and describe the fix.
- Contrast and text sizing. Sale badges, muted helper text and overlaid captions on hero imagery are the usual failures. This is measurable, so measure it rather than judging it.
- Images that say something. Decorative images take an empty alt. Product images do not. Alt text that repeats the filename is a finding.
- Audit what you bolted on. Review apps, the cookie banner and the announcement bar as part of the storefront, because a plaintiff's screening tool does not care which vendor shipped the component.
- Keep it from regressing. Automated checks in CI catch perhaps a third of issues, which is worth having, but only a manual keyboard and screen reader pass on the purchase path tells you whether the store is usable.
A realistic target is WCAG 2.1 AA on the paths that matter most, documented, with a dated remediation plan for the rest. A written plan with real dates has repeatedly been worth more in a negotiation than a widget subscription.
Where to start this week
Do the keyboard test yourself. Ten minutes, no tools, no budget. Put the mouse away and buy something from your own store. Most people do not finish.
Find out what you are running. If a theme was bought and customised, or apps were added over several years, nobody currently knows the state of the storefront. That is the actual problem.
Get the purchase path to AA first. Home, collection, product, cart, checkout. That is where complaints are written and where the revenue is.
Write the plan down and date it. Being mid-remediation with evidence is a materially different position from having done nothing.
This is the work we do. Smart-UI builds and supports Shopify stores, and accessibility to WCAG 2.1 AA is part of the storefront rather than a separate project. On Completedworks it was done alongside the Online Store 2.0 migration and the redesign, because the templates were being rewritten anyway and that is the cheapest time to fix it.
If you do not know where your store stands, the fixed-price audit covers accessibility alongside performance, SEO and security, and comes back as ranked findings with reproduction steps rather than a score. For Shopify specifically, the storefront work is quoted within 48 hours of a brief.
If you have had a demand letter, talk to a lawyer first, then send us the store and we will tell you what the engineering side of the response looks like.
Sources: EcomBack H1 2026 report, EcomBack Q1 2026, FTC order coverage, Seyfarth ADA Title III blog, Lainey Feingold on the FTC action, DOJ Title II rule analysis, Ogletree, DOJ deadline extension, Jackson Lewis, ABA Business Law Today on Title III.