The Legal Risk of an Inaccessible Website, Briefly

I generally avoid selling accessibility through fear, because fear produces the worst version of the work: the grudging minimum, done to make a threat go away. But people keep asking me about the legal side, and pretending it doesn’t exist would be its own kind of dishonesty. So here is the shape of it, as briefly and plainly as I can manage, followed by the more interesting question of what to do about it.

The legal landscape works roughly like this. In the United States, the Americans with Disabilities Act requires that places serving the public be accessible to people with disabilities. The law predates the commercial web, so it never mentions websites, and courts have spent years working out how it applies to them. The direction of travel, though, has been consistent: if your website is how the public does business with you, being effectively unusable by blind or motor-impaired customers looks a lot like a locked front door. In Europe, the situation is more explicit. The European Accessibility Act, in force for new consumer-facing products and services since mid-2025, directly requires accessibility from a wide range of businesses, including e-commerce, and it applies to companies outside Europe that sell in. Other countries have their own versions. The details differ; the trajectory doesn’t. Every year, more of the world treats web accessibility as an obligation rather than a courtesy.

There’s no law called WCAG, but in practice the Web Content Accessibility Guidelines are what everyone points to when they need a definition of “accessible.” Regulators reference them, settlements reference them, courts reference them. So while the statutes are vague about websites, the working standard is not, and it’s public, versioned, and testable. This is actually good news. Most legal risk is fuzzy. This one comes with a checklist.

What does the risk look like concretely? Demand letters and lawsuits over inaccessible websites are common enough in the US that entire practices exist on both sides of them, and the targets are not only giant companies. Small e-commerce sites get letters too, often generated from automated scans of exactly the kind of defect a machine can find: images without text alternatives, forms without labels, buttons that can’t be reached by keyboard. I won’t dress this up with statistics, because the numbers vary by year and by who’s counting, and because the precise count doesn’t change the decision you should make. It’s enough to know the letters exist, they arrive without warning, and responding to one costs real money even when it goes nowhere.

Here’s the economic asymmetry that I think should actually drive the decision. Once a letter arrives, everything is expensive. Lawyers are expensive. Settlements are expensive. Remediation done in a panic, under a deadline, by whichever contractor can start Monday, is the most expensive of all, and it tends to produce the bolted-on kind of fix, an overlay or a hasty patch, that often doesn’t hold up and doesn’t help actual users much. The same work done calmly, before anyone is angry, costs a fraction as much. Most of what these complaints cite is not exotic. Text alternatives on images, labels on form fields, sufficient color contrast, keyboard operability, focus that you can see. These are the basics of WCAG’s A and AA levels, and a competent developer can clear most of them incrementally, as part of normal work. The choice isn’t between spending and not spending. It’s between spending a little now on your own schedule, or more later on someone else’s.

And that reframing points at what I think is the honest conclusion. The legal risk is real, but it’s a lagging indicator. Lawsuits happen because real people hit real walls: a customer who couldn’t complete checkout, couldn’t read the menu, couldn’t book the appointment. Every legal complaint is preceded by some number of silent failures, people who didn’t sue, didn’t complain, and simply took their money elsewhere. If you fix your site only to the point where you’re unlikely to be sued, you’ve treated the symptom. If you fix it so that disabled customers can actually buy from you, the legal risk mostly evaporates as a side effect, and you’ve also gained customers. The defensive framing and the commercial framing prescribe nearly identical work; only the motivation differs, and motivation shows in the quality.

So my practical advice is anticlimactic. Don’t buy an overlay widget and call it done; the consensus among accessibility practitioners, and the experience of companies that tried, is that overlays don’t reliably fix the underlying problems. Don’t wait for a letter to find out what’s wrong. Find out yourself, now, while it’s cheap. Run an audit, automated first to catch the machine-findable majority, then a human pass over your critical flows. Fix things in order of how badly they block a real customer. I built GazeSite partly because this first step, just seeing your own defects, was harder than it should be. But whatever tool or consultant you use, the strategy is the same: the best legal position is a website that works. Everything else is paperwork about a door that should simply be unlocked.