PDF/A Explained: What Courts and Government Portals Want

PDF/A is the archival version of PDF, standardised as ISO 19005, that forces a file to carry everything it needs to look the same in decades: embedded fonts, no encryption, no scripts and no outside dependencies. Courts and archives ask for it because they must keep filings readable long after the software that made them is gone.

If an e-filing portal has rejected your document with a message about PDF/A, the file probably opened fine on your computer. That is the point. Ordinary PDFs can lean on fonts installed on your machine, links to outside content or features that a future reader may not support. PDF/A rules those shortcuts out. This guide explains the versions and levels, shows real court and government requirements, and walks through converting and checking a file.

Key takeaways

• PDF/A comes in four parts: PDF/A-1 (2005), PDF/A-2 (2011), PDF/A-3 (2012) and PDF/A-4 (2020), each published as part of ISO 19005.
• All fonts must be embedded, and encryption, JavaScript, audio and video are not allowed.
• Level b ("basic") preserves how pages look; level a adds tagged structure; level u adds Unicode text. PDF/A-4 drops these levels.
• Requirements differ court by court. The Delhi High Court names PDF/A as its preferred e-filing format, and some US federal courts have required it.
• Siliph's PDF to PDF/A produces PDF/A-2b. Check the result with a validator such as veraPDF if your court is strict.

What makes a PDF a PDF/A

The Library of Congress summarises the core rules in its format description (Library of Congress, PDF/A Family):

• "All fonts must be embedded and also must be legally embeddable for unlimited, universal rendering."
• "Encryption is disallowed."
• "Javascript and executable file launches are prohibited."
• "Audio and video content are forbidden."
• "PDF/A is constrained to avoid external dependencies."

Each rule removes a way a document could break later. A missing font turns text into substitutes or boxes. A password can be lost. A script may not run in tomorrow's reader. A link to an outside file may point nowhere in ten years.

Versions and conformance levels

The years come from the Library of Congress listing above. Within PDF/A-1, -2 and -3, a letter shows the conformance level. Using PDF/A-2 as the example, the Library of Congress explains that level b "indicates minimal compliance" designed to "ensure that the rendered visual appearance of a conforming file is preservable over the long term"; level u adds the requirement that "all text in the document have Unicode equivalents"; and level a "indicates complete compliance", which includes tagged document structure (Library of Congress, PDF/A-2a). PDF/A-4 "drops the three conformance levels (A, B, U)" and introduces profiles including PDF/A-4f and PDF/A-4e for engineering documents (Library of Congress).

For most filings, PDF/A-1b or PDF/A-2b is what a portal means when it simply says "PDF/A". If a rule names one part and level specifically, follow it exactly.

Who asks for PDF/A

Requirements are local, so always read your own court's rules. These examples show the range.

Indian High Courts

The e-Filing Rules of the High Court of Delhi say documents "should be converted into Optical Character Recognition (OCR) searchable Portable Document Format (PDF) or PDF/A" and that "PDF/A is the preferred format." Scanned documents should use "an image resolution of 300 DPI" in OCR searchable mode (Delhi High Court e-Filing Rules, Rules 6.2 and 6.3). The Rajasthan High Court's e-filing guidelines also call PDF/A the preferred format (Rajasthan High Court).

United States courts

There is no single national rule. The Western District of Pennsylvania's ECF manual stated that "Only documents in PDF/A format may be filed with the court using its ECF System" from January 2012 (W.D. Pa.). Other districts accept both: the Eastern District of Oklahoma says "CM/ECF accepts both PDF and PDF/A documents" (E.D. Okla.). In Florida, a 2019 administrative order made PDF/A "The preferred format for documents filed through the Florida Courts E-Filing Portal", while still allowing other formats (The Florida Bar News, 2019). Florida's rules have since been revised, so check the current version.

National archives

The US National Archives lists PDF/A-1 and PDF/A-2 as preferred formats for born-digital textual records, with ordinary PDF versions as acceptable (NARA, Appendix A: Tables of File Formats).

How to convert a PDF to PDF/A

Siliph's PDF to PDF/A rewrites a PDF to the PDF/A-2b rules: every font used is embedded, colours are converted to RGB with an sRGB colour profile attached, and the file is marked as PDF/A. It is free, needs no account and takes up to 50 PDFs per task, each up to 200 MB.

1. Remove any password first. PDF/A does not allow encryption, so a locked PDF must be opened with Unlock PDF before conversion. You need the password to do this.
2. Make scans searchable. A PDF/A made from a scan is still an image. If your court asks for OCR-searchable files, as Delhi does, run OCR PDF first. Our guide on making a PDF searchable covers scan quality.
3. Convert. Upload the file to PDF to PDF/A and click convert. There is no level to choose; the output is always PDF/A-2b.
4. Download and check. The file keeps its name with -pdfa added, so brief.pdf becomes brief-pdfa.pdf. Open it and confirm every page looks right.
5. Validate if it matters. For a strict portal, run the file through a validator before filing.

Two behaviours to know. Content that PDF/A does not allow is left out rather than stopping the job, so check that nothing you need has vanished. And colours are converted to RGB, which can shift print-specific CMYK shades slightly. Neither matters for a typical brief or affidavit.

If your court specifically names PDF/A-1b or PDF/A-3, a PDF/A-2b file may still be refused, because those are different profiles. Check the court's guidance or use a desktop tool that lets you choose the part.

How to check that a file really is PDF/A

A file can be labelled PDF/A and still fail the rules, for example if it was edited after conversion. The industry tool for checking is veraPDF, described by its developers as "a definitive, open source PDF/A validator", backed by the PDF Association and the Open Preservation Foundation (veraPDF). It is free to download and use.

Simple checks you can do without software:

• Some PDF readers show a notice when a file claims PDF/A conformance. A claim is not proof, so validate when it matters.
• Try to search for a word. If the original was a scan and search finds nothing, you still need OCR.
• Confirm there is no password prompt.

Common problems before filing

If size is the problem, Compress PDF can reduce a large scan, but re-check the PDF/A status afterwards, since any processing step can affect conformance. The order that works best is: unlock, OCR, compress, then convert to PDF/A last.

PDF to PDF/A runs on Siliph's servers. The file travels over an encrypted upload, is converted automatically with no human access, and is deleted within 2 hours along with the PDF/A copy.

Frequently asked questions

Is PDF/A the same as a normal PDF?

PDF/A is a restricted version of PDF. Any PDF reader can open it, but the file must follow extra rules, such as embedding all fonts and avoiding encryption, scripts and external links, so it displays the same in the future.

Which PDF/A level should I use for court filing?

Use the level your court names. If the rules only say PDF/A, PDF/A-1b or PDF/A-2b is what most systems expect, because the b level preserves the visual appearance of each page.

Can a PDF/A file be password-protected?

No. Encryption is not allowed in PDF/A. Remove any password before converting, and rely on the portal's own security for confidential filings.

Does converting to PDF/A make a scan searchable?

No. PDF/A preserves appearance; it does not recognise text. Run OCR on a scanned document first, then convert the searchable result to PDF/A.

Why is my PDF/A file larger than the original?

PDF/A requires every font used to be embedded in the file, and that adds data. The increase is usually small for text documents. Large scans are better compressed before the PDF/A conversion.