PDF/UA (ISO 14289) is the international standard for accessible PDF documents. A PDF/UA-conformant file has a correct tag tree, logical reading order, real table and form structure, meaningful text alternatives for images, and proper language and metadata — so a screen reader and other assistive technology can navigate it. It's the document equivalent of meeting WCAG for a website.

Isn't running a PDF through Acrobat's auto-tag enough?

Not for real-world documents. Acrobat's auto-tag is a starting point that can get close on simple text pages, but for complex tables, multi-column layouts, forms, or figures it produces a broken tag tree that a human still has to fix by hand. Genuine PDF/UA conformance for complex documents is manual, careful work — which is why quality supply is thin.

What laws require accessible PDFs?

Section 508 (US federal agencies and vendors selling to them), the EU's European Accessibility Act and EN 301 549, and HHS's Section 504 rule for federally-funded healthcare all apply to the documents an organization publishes — not just its website. Government, legal, finance, insurance, and education organizations with large PDF back-catalogs are especially exposed.

How much does PDF remediation cost?

Document remediation is priced per page, typically $1–$11.50 and clustering around $5–$8 per page for standard documents. Automated-only approaches are cheaper but unreliable for complex content. Large back-catalogs (tens of thousands of pages) are substantial one-time projects, and because organizations keep publishing new documents, ongoing per-document remediation is often a recurring engagement.

Can PDF remediation be automated?

Partially. Fully manual remediation is the norm, but a semi-automated pipeline — automated tagging plus reading-order heuristics plus AI-assisted alt-text drafting, with human QA on tables and figures — can deliver genuinely PDF/UA-conformant files faster without the corners that automation-only tools cut. This is where an engineering team has a real cost advantage over manual-only document shops.

Accessibility

Accessible PDFs (PDF/UA): What Document Remediation Actually Involves

Section 508, the EAA, and HHS rules all require accessible documents — not just websites. Here's what PDF/UA remediation is, why 'run it through Acrobat' isn't enough, and what it costs.

July 28, 20262 min readAccessibility

Short answer: PDF/UA (ISO 14289) is the standard for accessible PDF documents. Remediating a PDF means giving it a correct tag tree, logical reading order, real table and form structure, and text alternatives — so a screen reader can navigate it. "Running it through Acrobat's auto-tag" is not enough for complex documents, which is exactly why this work is in demand and under-supplied. If a government, legal, finance, or education buyer requires accessible documents, here's what's actually involved.

Why Documents, Not Just Websites

Accessibility law doesn't stop at your homepage. Section 508 (US federal and anyone selling to them), the EU's EAA and EN 301 549, and HHS's Section 504 rule all apply to the documents an organization publishes — benefits statements, court filings, tax forms, financial disclosures, medical forms, contracts, and RFPs. Organizations in these sectors sit on back-catalogs of *tens of thousands* of legacy PDFs, every one of which is in scope.

What "Remediation" Actually Means

A PDF looks fine to a sighted reader while being completely opaque to a screen reader. Making it conformant to PDF/UA involves:

Tag tree — the hidden structure that tells assistive tech what's a heading, paragraph, list, or figure.
Reading order — the sequence a screen reader follows, which often differs from the visual layout in multi-column pages.
Tables — real header/data cell associations, not just visually-aligned text.
Forms — labelled, keyboard-operable fields with clear instructions.
Alt text — meaningful descriptions for images, charts, and figures.
Language, bookmarks, and metadata — so navigation and pronunciation work.

Why "Just Use Acrobat" Fails

Acrobat's auto-tag feature is a starting point, not a solution. For simple text documents it can get close; for anything with complex tables, multi-column layouts, forms, or figures — which is most real-world documents — it produces a broken tag tree that a human still has to fix by hand. This is why credible PDF remediation is manual, tedious work, and why supply is thin: it's boring, so few do it well.

What It Costs

Document remediation is priced per page and the market is fairly transparent: roughly $1–$11.50 per page, clustering around $5–$8 for typical documents, with automated-only approaches cheaper but far less reliable for complex content. A single government or enterprise back-catalog of 20,000 pages is a substantial, and often recurring, engagement — because organizations never stop producing new documents.

Where Engineering Changes the Economics

This is the one accessibility niche where an engineering team has a real cost advantage. Most vendors remediate PDFs entirely by hand. A senior team can build a semi-automated pipeline — automated tagging plus reading-order heuristics plus AI-assisted alt-text drafting, with human QA on the hard parts (tables, figures) — that delivers genuinely PDF/UA-conformant files faster, without the corners automation-only tools cut. Our background building document-intelligence and RAG pipelines is directly what makes that possible.

We remediate document back-catalogs to PDF/UA and keep new documents conformant as they're published — see Accessibility Audit & WCAG Remediation, or start with a free gap read.