# Overlay Widgets Relocate Accessibility Risk Rather Than Reducing It ## Claim An accessibility overlay sits on top of existing barriers instead of removing them, so it changes what a site looks like it is doing about access without changing what the underlying code does. <!-- atomicity: 1 idea ✓ --> ## Scope & conditions - Holds for script-injected overlays and toolbars marketed as compliance solutions, where the remediation happens at runtime in the client rather than in the source markup. - Breaks for tooling that edits the actual source — a scanner that opens a pull request against your markup is not an overlay, whatever it is sold as. - The legal exposure argument is jurisdiction-bound and currently US-shaped; the mechanical argument (broken code stays broken) is not. - Says nothing about whether a given overlay helps an individual user in a given moment. It argues the overlay does not discharge the obligation. ## Evidence > [!quote] > Because widgets don't fix underlying accessibility issues. They don't repair broken code, label unlabeled buttons, or ensure proper navigation. Vanessa Tillemans states the mechanism directly and pairs it with claimed litigation data: roughly one in four ADA website suits in 2023 targeted sites that already had a widget installed, with AccessiBe, UserWay, and AudioEye together named in 929 such suits, plus a 2024 FTC action against AccessiBe over its compliance marketing. Those figures are her citations rather than independently verified here, and should be checked against UsableNet and the FTC before being repeated publicly. The mechanical claim stands independently of the numbers: an overlay that cannot reach the DOM's underlying semantics cannot repair them. Joe Dolson names the reputational half of the same problem from the other direction — practitioners hear "accessibility plugin" and assume overlay, which is why a canonical WordPress plugin in this space has a naming problem before it has a code problem. ## Links - [[3. Resources/Permanent Notes/Accessibility Remediation Inside Vendor Tooling Trades Speed for Client Autonomy]] — because extends; that note is about maintainability risk, this one about whether the barrier is removed at all - [[3. Resources/Permanent Notes/A Passing Automated Accessibility Score Is a Floor Not a Verdict]] — both describe a signal that reads as compliance while the underlying failure persists - [[3. Resources/Permanent Notes/A Vendor Accessibility Report Is Disclosure Not Certification]] — because contrasts on what a vendor artifact can honestly claim - [[3. Resources/MOCs/Accessibility & Disability Inclusion MOC]] — belongs with "making the case commercially," as the counter-case ## Tests / falsifiers - Pull the UsableNet dataset directly and compute the widget-installed share of filed suits year over year. If widget-using sites are sued at or below base rate, the exposure claim fails while the mechanical one survives. - Run an automated scan plus manual AT testing on a site before and after installing an overlay, diffing defects at the source level. Real source-level improvement would falsify the "relocates rather than removes" framing. ## Maintenance - Last tended: 2026-09-08 - Next review: 2026-12-08 ## Source - [[3. Resources/Readwise/Articles/Accessibility Lawsuits and WordPress The Violations Every Business Needs to Know]] — Vanessa Tillemans, WordPress Accessibility Day 2025. - [[3. Resources/Readwise/Articles/All About Accessibility Lab With Joe and Anne Save the World Brewing TX Cerveza]] — Joe Dolson on the naming risk.