Methodology
A reproducible five-page sampling protocol
The starting frame is the entire .it subset of Tranco list 46XZX (8 July 2026), which matches the original domain cohort used for the Italian Web Security Study (IWSS). Collection took place from 9 to 13 October 2026.
Chromium pages were rendered with Playwright and evaluated with axe-core 4.14.0. Up to five pages per domain were selected using predetermined homepage, functional, content, deep-path and seeded rules. Automated tests were restricted to WCAG 2.0, 2.1 and 2.2 Level A and AA tags.
The Tranco sample is popularity-ranked and is not representative of all Italian websites.
Scope and limitations
Automated checks do not certify accessibility
Of 10,022 sampled domains, 2,449 were not evaluable, including 1,150 that returned HTTP 403. These domains were not classified as either accessible or inaccessible. Among evaluable domains, 7,328 had at least one detected violation (96.8%).
The browser runs used a fixed desktop viewport, without sign-in, manual keyboard inspection or assistive technology. Excluding 470 cross-domain redirects yields a similar observed prevalence of 96.7%.
A manual stratified review and an identified page-manifest anomaly were still awaiting final disposition at the time of manuscript preparation.
