Accessibility ≠ Overlay: an automated tool does not replace a real audit
Automated tools can find certain accessibility issues quickly, repeatedly and at scale. W3C says they can save time and effort, but is also explicit: no tool alone can determine whether a site meets accessibility standards. Knowledgeable human evaluation is required.
What a tool cannot decide
Some requirements depend on meaning and context. A tool may verify that an image has an alt attribute; deciding whether the text communicates what is needed can require human judgement. An automated result also does not prove that a person can complete a purchase, understand an error, operate a dialog by keyboard or navigate with a screen reader. W3C also warns that tools can produce inaccurate or misleading results.
An audit is more than clicking “scan”
A serious evaluation defines scope, explores the site, selects representative pages and states, performs checks and documents findings. WCAG-EM structures this type of process. W3C recommends evaluating early and throughout development, combining tools, manual review and testing. Involving people with disabilities can add evidence about real user experience.
Overlays, Blind Reader and the essential distinction
Automating the detection or modification of certain patterns is not the same as understanding a person’s complete experience. An overlay therefore does not demonstrate that the whole website is accessible. Blind Reader, as assistive technology, can help the user access or work with content, but it does not turn an automated result into an audit or certify WCAG conformance. The solid combination is tools + human knowledge + real testing. Delivery 14 will address why an automated “100/100” is not WCAG conformance.