Many people first hear about PAC after receiving an email from a client: the delivered PDF failed an accessibility check, does not meet the PDF/UA standard, and must be fixed before it can be submitted again.
The challenge is that the email usually includes a report listing hundreds—or even thousands—of issues. Most document production teams have no idea where to begin.
Why does a PAC check matter?
PAC (PDF Accessibility Checker) is an accessibility checking tool designed to assess whether a PDF meets PDF/UA requirements.
People with visual impairments cannot see a carefully designed layout. They rely on screen readers to read content aloud, line by line. When a PDF lacks proper structural tags, the screen reader may read the document in a confusing way:
- It cannot distinguish heading levels from body text;
- When reading a table, it cannot associate data cells with their corresponding headers;
- When it encounters a hyperlink, it may mechanically read out a long URL;
- It may even read background color blocks and decorative lines as body content, making information difficult to access.
The core purpose of a PAC check is to verify that a file is genuinely usable for people who rely on assistive technology. Documents that look visually polished can still contain major structural gaps from PAC's perspective.
That is why government, education, healthcare, and financial organizations increasingly treat PAC checks as a firm delivery requirement. If a file cannot pass, it may be returned or prevented from publication.
Why do PDFs exported from InDesign or Word often fail compliance checks?
When exporting a PDF from InDesign, saving a file as PDF from Word, or printing a PDF through Acrobat, most teams focus on whether the layout looks right—not whether the underlying structure is correct.
Below are several common PAC errors and the reasons behind them:
1. Decorative page elements become accessibility noise
Common PAC error:
Path object not taggedorContent shall be marked as Artifact
Divider lines, background color blocks, and decorative borders are visual elements that do not need to be read aloud. If they are not marked as decorative artifacts during export, PAC interprets them as undefined content and reports an error. A technical manual can contain hundreds or thousands of these lines and blocks, quickly creating a report full of issues.
2. Screen readers cannot tell table rows from columns
Common PAC error:
Table header cell has no scope attribute
Tables exported from InDesign often lack the Scope attribute for header cells. As a result, screen readers cannot establish the relationship between data cells and headers. Instead, they read one cell after another without conveying the table's logical meaning.
3. Links are read as long URLs
Common PAC error:
Link annotation does not have an alternate description
By default, a PDF hyperlink may use its URL as its accessibility description. When a screen reader reaches the link, it reads the full address aloud, such as https://..., which interrupts the reading experience. PAC requires every link to have a clear text description, or it will report the issue.
4. The file lacks metadata that identifies PDF/UA compliance
Common PAC error:
pdfuaid:part is missing
A PDF's underlying information, known as XMP metadata, needs a dedicated PDF/UA version identifier to declare that the document meets the accessibility standard. Many tools, including some versions of InDesign, do not automatically write this identifier during export, causing the file to fail a basic check.
Why is it difficult to fix these PDF/UA errors manually?
After receiving a PAC report, the most frustrating part is that it lists rule IDs and error locations without explaining how to fix them.
With a fully manual workflow, you need to open Acrobat's Tags panel and edit the PDF's underlying structure almost like editing code. For document production and translation teams, that can be a significant technical hurdle.
A 50-page technical manual can easily produce more than a thousand errors. Even if you know how to address each issue, fixing them one by one manually can take an impractical amount of time.
SimplifyAI's automated remediation workflow
For this type of repetitive work, SimplifyAI provides an automated workflow that combines machine processing with human confirmation.
Upload the document, and the system runs a complete check and organizes large volumes of issues by category:
- Issues the system can handle: For example, batch-marking decorative graphics or inferring and completing table header scope. The system generates a remediation plan automatically.
- Issues that need human confirmation: For example, descriptions for images and hyperlinks. The system extracts a default description for links, which you can review and edit directly in the interface. For images, you can enter a description or choose to mark them all as decorative images.
After you confirm the approach, select repair. Once the file is updated, the system runs another PAC check and delivers a before-and-after remediation report together with the final PDF.
Which issues can be fixed automatically?
Fully automated processing:
- Clean up decorative elements: Automatically mark backgrounds, lines, and other visual elements as artifacts so they do not interrupt screen reader output.
- Repair tables: Automatically infer and complete row or column scope attributes for table headers (TH Scope).
- Add compliance identification: Automatically add PDF/UA compliance information to the file's XMP metadata.
Issues that need a quick human addition or review:
- Hyperlink descriptions (Contents): The system extracts visible link text or the target address as default screen-reader content. You can review and revise it directly to support a better reading experience.
- Alternative text for images and formulas (Alt text): Purely decorative images can be batch-marked as decorative with one action. For meaningful images, users can enter a text description directly in the interface.
Issues that cannot currently be fixed automatically:
- Source font issues (incomplete CIDSet): These issues originate in the font embedding settings used when the PDF was exported. Forcing a fix may corrupt text, so the system flags a warning in the report and recommends asking the document production team to re-export from the source file.
Common use cases
- Government, financial, healthcare, and other industries that require PDF/UA-compliant files
- Technical documents, user manuals, and product guides that must pass a client's PAC check
- Batch remediation of existing PDF archives to meet accessibility requirements
If your PDF needs to pass a PAC check, upload it to run a free initial check first. You can review the issue distribution and estimated remediation effort before deciding whether to proceed.