Conference 26–27 September 2022 📍 Washington D.C., USA

PDF Days North America 2022 — The Return to In-Person After COVID

PDF Days North America 2022 took place in Washington D.C. on 26–27 September 2022 — the first in-person PDF Association conference in North America since the pandemic. Sessions covered PDF 2.0 implementation status, accessibility mandates under US Section 508 and the ADA, digital signature trust frameworks, and the PDF Association's new specification clarification process.

Official event page →

Key takeaways

  • PDF 2.0 was 5 years old but real-world adoption of its new features (AES-GCM, new structure elements, page-level metadata) remained limited — conservatism driven by installed-base compatibility concerns.
  • Section 508 enforcement: PDF accessibility was consistently among the top OCR complaint categories for US federal agencies — primarily inaccessible PDFs on government websites.
  • Auto-remediation tools had matured by 2022 but full automation remained beyond reach for complex tables, multi-column layouts, and figure captions.
  • PDF/UA-2 development previewed: /B conformance level removal, normative MathML, WCAG 2.1 alignment — published April 2024.
  • Federal agency legacy PDF remediation: triage approach adopted — prioritise frequently accessed documents, provide HTML alternatives for highest-traffic legacy content.
PDF Days North America 2022 was the first in-person North American PDF conference since the COVID-19 pandemic forced the 2020 and 2021 editions online. Held in Washington D.C. in October 2022, the conference marked a return to face-to-face engagement for the North American PDF community — developers, document engineers, government digital transformation teams, and accessibility professionals. The event arrived at a significant moment: ISO 32000-2:2017 (PDF 2.0) was five years old, yet real-world adoption of new PDF 2.0 features remained limited. Sessions addressed this implementation gap honestly, examining why PDF 2.0 feature adoption lagged the specification timeline and what organisational and tooling factors were responsible. For the accessibility community, the conference came at a pivotal point in US federal accessibility enforcement — Section 508 compliance was under increasing scrutiny, and the new PDF/UA-2 specification was in advanced development, promising the first update to the PDF accessibility standard in a decade.

PDF 2.0 at five years: implementation status report

Five years after publication, sessions at PDF Days North America 2022 assessed the real-world state of PDF 2.0 adoption: Widely implemented features — Cross-reference streams (present since PDF 1.5, standard in all modern generators), object streams, and basic encrypted PDFs were near-universal. AES-256 encryption was widely supported. These features predated PDF 2.0 or were straightforward to implement. Slow-to-adopt features — The genuinely new capabilities of PDF 2.0 were seeing limited adoption: - AES-256-GCM (authenticated encryption) — Few generators produced GCM-encrypted PDFs; most still used CBC mode from PDF 1.7 - Unencrypted wrapper documents — Almost no implementations - Page-level /Metadata — Inconsistent support in both generators and parsers - New structure elements (/Aside, /FENote, /DocumentFragment) — Not produced by any major authoring tool Reasons for slow adoption — The PDF format's installed base creates a strong conservatism pull: generating features that older viewers might mishandle is a support risk. The PDF 2.0 specification had also had significant errata in its first years, creating uncertainty about which version of the spec to implement. The ISO 32000-2:2020 corrigendum addressed many ambiguities but added to the perception of instability. Practical takeaway — For document producers and tool developers: target PDF 2.0 for new features when supported by your toolchain, but do not assume recipients have PDF 2.0-capable viewers for anything beyond basic features. See PDF Version History for context.

Section 508 accessibility enforcement in US federal agencies

Section 508 of the US Rehabilitation Act (as amended by the Workforce Innovation and Opportunity Act) requires federal agencies to make electronic information and communication technology accessible. For PDFs, this primarily means meeting the PDF/UA accessibility standard (ISO 14289-1). Sessions at PDF Days North America 2022 reviewed the enforcement landscape: Access Board role — The US Access Board publishes the Section 508 standards. The 2017 refresh (final rule) explicitly referenced WCAG 2.0 and incorporated PDF/UA requirements for PDF documents. The 2022 conference reviewed enforcement action patterns since the refresh. OCR complaint data — The US Department of Justice's Office of Civil Rights tracks 508 complaints. PDF accessibility was consistently among the top complaint categories — primarily from federal agency websites delivering inaccessible PDF documents to the public. Government benefit documents, tax forms, and agency publications were the most commonly reported categories. Auto-remediation tools — maturity and limits — By 2022, auto-tagging and remediation tools (Adobe Acrobat Pro's accessibility wizard, axesPDF, CommonLook, Foxit's accessibility features) had matured significantly. Agencies that had implemented automated accessibility checking in document publishing workflows were reporting meaningful improvement in compliance rates. However, complex tables, multi-column layouts, and figure captions continued to require manual remediation — full automation remained beyond current tool capabilities. The document legacy problem — Federal agency websites contain millions of legacy PDFs produced before accessibility requirements applied. Complete remediation of legacy content is not practically feasible; agencies were developing triage policies: prioritise frequently accessed documents, automate what can be automated, provide accessible alternatives (HTML pages) for the highest-traffic legacy PDFs.

PDF/UA-2 development: preview of the 2024 standard

At the time of PDF Days North America 2022, PDF/UA-2 (ISO 14289-2) was in advanced development. Sessions at the conference previewed the most significant planned changes: Removal of /B conformance level — PDF/UA-1 had a "basic" (/B) conformance level that allowed a reduced requirement set. PDF/UA-2 would eliminate this level — there would be only one conformance level, and all requirements would be normative. This was controversial at the conference: organisations that had invested in /B compliance would need to do additional work. MathML requirement — The planned normative requirement for mathematical content to be expressed in MathML was previewed. STEM publishers at the conference raised implementation timeline concerns — LaTeX-to-MathML workflows existed but were not yet mainstream for production publishing. Alignment with WCAG 2.1 — PDF/UA-2 would align with WCAG 2.1 rather than WCAG 2.0. The additional WCAG 2.1 success criteria — particularly non-text contrast (1.4.11) and text spacing (1.4.12) — would create new requirements for PDF document design. PDF/UA-2 was ultimately published in April 2024 as ISO 14289-2:2024. See PDF/UA-2 Complete Guide for the full details of the published standard.

Frequently asked questions

What does Section 508 require for PDF documents?

Section 508 of the US Rehabilitation Act (as updated in the 2017 refresh, applying WCAG 2.0 Level AA criteria) requires that PDF documents accessible to the public via federal agency websites or applications must meet WCAG 2.0 Level AA success criteria. For PDFs, this is operationalised through PDF/UA (ISO 14289-1), which defines the PDF-specific implementation requirements for WCAG success criteria. In practice: PDFs must be tagged, have alternative text for images, declare the document language, provide a title, have reading order reflected in the tag tree, and not use security settings that block assistive technologies. The Access Board's WCAG guidance at section508.gov provides agency implementation guidance.

Did PDF 2.0 break backward compatibility with PDF 1.7 viewers?

No — PDF 2.0 is designed for backward compatibility. A PDF 2.0 document will open in a PDF 1.7 viewer, which will ignore features it doesn't understand. The PDF specification includes this tolerance by design: unknown keys in dictionaries are ignored; unrecognised operators in content streams cause only that operation to be skipped. The practical consequence: a PDF 2.0 document using only PDF 2.0-specific features (like AES-256-GCM encryption or new structure elements) will display correctly in an older viewer but may: not decrypt correctly (if the viewer doesn't support GCM), show simplified rendering (if new structure elements are relied upon for layout), or miss accessibility features (if the viewer doesn't process new tag types). The conservative approach is to use PDF 2.0-specific features only for enhancements that degrade gracefully — accessibility features, richer metadata — rather than for features essential to rendering.

Try it in your browser

The PDF tools and standards discussed at this event are available free — no account needed.