Briefing
The European Accessibility Act Reached Your Manuals

The Deadline That Already Passed
June 2025 Was the Start Line
The EAA's obligations applied from June 28, 2025, across a defined range of products and services sold into the EU, and with them, the digital content those products and services carry: instructions, support documents, customer-facing files. The measuring stick is EN 301 549, the European standard for accessible ICT, whose September 2026 revision moves documents from WCAG 2.1 to WCAG 2.2 Level AA: real headings, explicit reading order, text alternatives, semantic tables. It is not yet cited in the Official Journal under the EAA, so it is the yardstick rather than a legal safe harbor, but it is the one regulators and customers will measure against. Our guide to EN 301 549 V4.1.1 covers what changed for manuals.
For organizations whose customer content is a library of exported PDFs, this lands on a familiar bruise: those files are photographs of documents, and a photograph has no headings to declare.
The Fork
Remediate Forever, or Fix the Source

The reflex response is remediation: hire the taggers, fix the files. For a small, static set that's correct. For a living estate it's a treadmill: labor applied to output, re-applied at every revision, non-compliant again the moment one untagged export ships. The scalable answer is upstream, and it's the argument this site keeps making with a legal deadline now attached: a screen reader is a machine reader, and machine readers are served by structure, not paint. Content whose headings, order, alternatives, and tables are real at the source renders to accessible HTML and properly tagged PDF through a single-source pipeline: conformance achieved once, in the model and the template, inherited by everything after.
The practical sequence mirrors every estate problem we've written about: serve the library through a platform that can offer an accessible rendition alongside the PDF, move the living document families to structured authoring, and let remediation handle only the static tail. Firms selling into the EU have a compliance clock running either way. The choice is whether each tick costs a tagging invoice, or nothing, because the structure was already true.
FAQ
Questions We Hear
What is the European Accessibility Act, and when did it apply?
An EU directive setting accessibility requirements for a defined range of products and services sold in the EU: consumer devices and terminals, e-commerce, banking, e-books, transport services, and more. Its obligations have applied since June 28, 2025, with a transition to June 28, 2030 for products already in use to provide services. For documentation teams the key point is scope: Annex I requires instructions for use to be publicly available, available through more than one sensory channel, in text formats that can generate assistive formats, and with alternatives for non-text content. The requirements reach the documents around in-scope products and services, not just the app's buttons.
What standard are documents measured against?
EN 301 549, the European standard for accessible ICT. Its current version, V4.1.1 (September 2026), aligns documents with WCAG 2.2 Level A and AA; the previous V3.2.1 used WCAG 2.1. No version is yet cited in the Official Journal under the EAA, which is the step that would make conformity a legal presumption, so today it is the reference point rather than a safe harbor. In practice it means real heading structure, reading order, text alternatives for images, sufficient contrast, machine-readable language, and tables that are tables. V4.1.1 also drops the earlier wording that let documentation be accessible in at least one of several formats.
Why doesn't per-document remediation scale?
Because remediation is manual labor applied to output. Tagging one PDF properly is a real job; doing it for a thousand-file library, and again for every revision of every file forever, is a treadmill, and the moment an untagged revision ships, you're non-compliant again. Remediation vendors are the right answer for a small static set and the wrong economics for a living document estate. The estate-scale answer is upstream: content whose structure is real at the source, so every rendition inherits accessibility instead of having it painted on.
How does structured content change the compliance math?
Structured content is accessibility's raw material: headings that are declared, reading order that is explicit, images with text alternatives as data, tables with semantics. Render that source to HTML and to tagged PDF through a proper pipeline and conformance is substantially a by-product: achieved once in the template and the model, inherited by every document and every revision after. It's the same one battle we keep describing: the structure that serves screen readers is the structure that serves search, integrations, and AI. Regulated accessibility just added a legal deadline to the argument.
Get In Touch
Off the Remediation Treadmill
If EAA compliance currently means a per-PDF invoice, the economics say fix the source. We build the structured pipelines where accessibility is a by-product, not a project.