Conference 21–22 September 2021 📍 Online

PDF Days North America 2021 — Virtual: Section 508 Enforcement, PDF/UA Progress, and Remote Signing Surge

PDF Days North America 2021 was held virtually on 21–22 September 2021. The event addressed the surge in electronic signing driven by remote work, Section 508 enforcement actions accelerated by pandemic-era government digitisation, and progress on PDF/UA-2 specification development. Digital signature interoperability between US and EU frameworks was a recurring theme.

Official event page →

Key takeaways

  • Two years of Section 508 enforcement had produced practical experience: complex tables and forms were the most common failure categories; automated tools were necessary but insufficient.
  • AI-assisted PDF tagging was emerging but not yet production-ready for high-stakes federal compliance; practitioner-led remediation tools had improved significantly.
  • US federal PDF 2.0 adoption was slow due to explicit version requirements in agency policies, NARA guidance gaps, and conservative court filing system requirements.
  • Procurement contract language requiring accessible PDF delivery and vendor remediation obligations was the most effective organisational lever for sustained compliance.
  • Screen reader testing with JAWS, NVDA, and VoiceOver remained essential alongside automated validation — tool-based checks could not confirm document usability.
PDF Days North America 2021 was held virtually in September 2021, organised by the PDF Association. While the European editions focused heavily on PDF 2.0 specification development and EU regulatory compliance, the North American editions were distinctly shaped by the US federal accessibility landscape. The September 2021 conference took place two years after the August 2019 deadline by which US federal agencies were required to have brought all ICT (including PDFs) into compliance with Section 508 of the Rehabilitation Act. Two years of lived experience with Section 508 enforcement, combined with the hybrid working transition, provided a rich empirical basis for sessions at the 2021 conference. Agencies that had gone through audit cycles and remediation campaigns could share what had worked, what had not, and what the automated testing landscape looked like in practice. The 2021 North American conference also addressed the US federal adoption of PDF 2.0, which was considerably more cautious than the European context — US federal document specifications often maintained explicit PDF version requirements, and the transition to PDF 2.0 in government workflows was slower than in commercial contexts.

Two years of Section 508 enforcement: lessons learned

The August 2019 deadline had been a forcing function, but Section 508 compliance for PDFs was understood as an ongoing operational requirement rather than a one-time project. By September 2021, US federal agencies and their contractors had two years of enforcement experience to draw on. Sessions at PDF Days North America 2021 reported on what that experience had revealed: Automated testing limitations — Automated tools like PAC, veraPDF, and the Acrobat Accessibility Checker could identify structural errors reliably — missing tags, improper reading order, tables lacking header associations — but could not assess semantic correctness. A document could pass automated checks and still be inaccessible to a screen reader user if the underlying content was poorly structured or the reading order did not reflect the visual layout. Table complexity — Complex tables — merged cells, multi-level headers, nested tables — remained the most reliably difficult content type to tag correctly and verify. Sessions covered PDF table tagging approaches and the specific table header association requirements in the W3C PDF techniques for WCAG 2.1. Forms and interactive elements — PDF forms with AcroForm fields had specific Section 508 requirements around tooltip text (accessible names), tab order, and label association. Sessions reported that forms were the second most common category of Section 508 failures after complex tables, particularly in legacy forms created before accessibility requirements existed. Procurement contract language — Agencies that had incorporated specific PDF accessibility requirements into procurement contracts — requiring vendors to deliver accessible PDFs and to provide remediation — were reporting better outcomes than agencies that had relied on general accessibility statements. Model contract language for PDF accessibility requirements was discussed.

Auto-tagging and remediation tool maturity

The remediation tool landscape had developed significantly between 2019 and 2021. Sessions at the 2021 conference reviewed the current state of automated tagging and AI-assisted remediation: Acrobat Pro's auto-tag feature — Adobe Acrobat Pro's auto-tagging had improved through several releases. The quality of auto-tagging on complex documents — those with multi-column layouts, tables, figures with captions, and mixed content types — had increased, but the fundamental limitation remained: auto-tags required human review and correction before a document could reliably pass accessibility validation or user testing with screen readers. CommonLook and axesPDF — Specialist remediation tools like CommonLook and axesPDF had developed semi-automated workflows that significantly reduced the manual effort required for full remediation. Sessions from practitioners using these tools in large-scale federal remediation projects discussed realistic throughput numbers and quality assurance approaches. AI-based tagging — Early-stage AI-assisted PDF tagging was emerging as a research and product development area. While not yet production-ready for high-stakes federal accessibility compliance, demonstrations showed that machine learning models trained on tagged PDF corpora could identify document structure elements with useful reliability on common document layouts. The expectation was that AI-assisted tagging would become a significant accelerator for remediation workflows within a few years — an expectation that proved accurate by 2024–2025. Testing with assistive technology — Sessions emphasised that tool-based validation was necessary but not sufficient. User testing with actual screen reader software (JAWS, NVDA, VoiceOver) was the only reliable way to assess whether a document was actually usable by people with disabilities. Guidance on lightweight testing protocols for organisations that could not run full user studies was a practical takeaway from these sessions.

PDF 2.0 in US federal workflows

US federal agencies' adoption of PDF 2.0 was significantly slower than in European commercial contexts, for several reasons discussed at the 2021 conference: Explicit version requirements — Many US federal document standards, procurement specifications, and agency IT policies specified PDF versions explicitly. The Federal CIO Council's document standards, agency records management policies, and court filing requirements (particularly in the federal court CM/ECF system) variously specified or implied PDF 1.4, PDF 1.5, or PDF 1.7. Agencies needed formal policy updates to permit PDF 2.0 output. Validation infrastructure — Agencies that had deployed PDF/A validation workflows (particularly for records management under NARA requirements) needed to update their validation configurations to handle PDF/A-3 and eventually PDF 2.0 documents. NARA guidance — The National Archives and Records Administration (NARA) had not yet issued updated guidance on PDF 2.0 for federal records. Since NARA guidance shapes what formats agencies can use for permanent records, this was a significant barrier to federal PDF 2.0 adoption. Sessions noted that NARA's historical pattern was to issue guidance several years after a format achieved broad commercial adoption. Court filing systems — The federal court CM/ECF (Case Management/Electronic Case Files) system had specific PDF requirements, and federal courts were conservative about format updates that might affect the integrity of judicial records. PDF 2.0 adoption in court filings was expected to lag commercial adoption by several years.

Frequently asked questions

What is Section 508 and how does it apply to PDFs?

Section 508 of the Rehabilitation Act requires US federal agencies to ensure that their ICT (Information and Communications Technology) — including electronic documents like PDFs — is accessible to people with disabilities. For PDFs, this means meeting the requirements of WCAG 2.0 Level AA as incorporated by the 2017 Section 508 Refresh (the revised standards published by the US Access Board). In practice, Section 508 compliance for PDFs requires: proper structure tags, correct reading order, meaningful alternative text for images, accessible table markup (header associations, scope attributes), accessible form fields (labels, tooltips), sufficient colour contrast, and no reliance on colour alone to convey information. The August 2019 transition deadline required all existing ICT — including legacy PDF libraries — to be brought into compliance, not just new documents.

What is CM/ECF and what are its PDF requirements?

CM/ECF (Case Management/Electronic Case Files) is the federal court system used by all US federal district courts, bankruptcy courts, and courts of appeals for electronic filing of legal documents. CM/ECF has specific PDF requirements: documents must be text-searchable (not image-only scans), must not be password-protected in ways that prevent downloading or printing, must comply with file size limits (typically 25MB or 50MB per document depending on the court), and must be filed as individual PDF files (not ZIP archives). Some courts have specific requirements about PDF version and feature usage. The transition to PDF 2.0 in federal court filings was expected to be slow given CM/ECF's conservative approach to format changes and the need to preserve multi-decade archival integrity of judicial records.

What NARA guidance exists for PDF in federal records management?

The National Archives and Records Administration (NARA) issues guidance on acceptable file formats for federal records through its Bulletin 2014-04 (Format Guidance for the Transfer of Permanent Electronic Records) and subsequent updates. PDF and PDF/A are listed as acceptable formats for permanent records. NARA's guidance has historically favoured PDF/A for archival permanent records due to its self-containment requirements (embedded fonts, no external dependencies, no JavaScript). PDF/A-1 and PDF/A-2 are specifically endorsed; PDF/A-3 is noted as acceptable with caveats about the embedded file relationship. NARA guidance typically lags several years behind commercial format adoption — agencies seeking to use PDF 2.0 for permanent records needed to check whether current NARA guidance explicitly addressed it or whether they needed to obtain NARA concurrence for non-standard format usage.

Try it in your browser

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