The short answer to that question is that all web content, including PDFs, must comply with the WCAG 2.2 standard at AA level. However, the problem with the WCAG standard in relation to PDF files is that “W” stands for “Web” and “G” for “Guidelines”, so it is both very HTML-oriented and, with guidelines, does not provide many options for physically testing PDF files.
This is where PDF/UA comes into its own, which is a technical ISO standard specifically for achieving accessibility in PDF files, so in a nutshell PDF/UA is the technical equivalent of WCAG for PDF, as they are both based on the same principles. That’s why we always recommend all public authorities to require PDF/UA compliance from tools and vendors, as it gives you the only real possibility to test PDF files afterwards and at the same time a long-lasting solution based on an ISO standard.
You CAN ensure even longer durability of your PDF files by also requiring them to comply with the PDF/A standard, which is an ISO standard for long-term archiving in PDF. There are several sub-variants of PDF/A, but if we concentrate on PDF/A-2, there are three sub-variants:
PDF/A-2b: “b” stands for “basic” and means that the PDF file must simply be a correct reproduction of the original document and in that format can easily consist of just scanned pages from a to z.
PDF/A-2u: As “b” variants, must be a correct reproduction of the original document, but text/lettering must also be included in unicode format.
PDF/A-2a: As “u”, must have text/text included in unicode format but also be a structured document (tagged).
If an accessible PDF file must also comply with PDF/A, it must therefore comply with the “a” variant, since accessibility is based on the logical structure.
So in summary:
- Available PDF files must ALWAYS comply with the PDF/UA standard
- Accessible PDF files MAY also comply with PDF/A-2a if there are no external dependencies
- PDF/A-2b can NEVER be used to measure accessibility in a PDF.