ShredHub
Back to blog
How-to7 min readSeptember 21, 2026

How to remove metadata from a PDF

Author, company, software and revision history travel with every PDF. Four ways to strip them — and how to verify the result.

A PDF looks final. Behind the pages, though, it carries a small record of how it was created: who authored it, which application produced it, when it was last modified, and sometimes the document ID. Send it to a client as-is and all of that goes along.

What you are actually removing

Fig. 01Two places hold the same story
Info dictionary/Info object
/Author/Creator/Producer/CreationDate/ModDate
XMP metadataXML packet
xmp:CreatorToolxmp:CreateDatexmpMM:DocumentID
Embedded objectscheck before sending
JavaScriptattachmentsform fields
The Info dictionary is the original PDF metadata. XMP is the newer, XML-based layer that can repeat and extend it. Cleaning only one leaves the other behind.PDF 32000-1:2008 and ISO 16684-1 (XMP)

A quick way to see what is inside is the command-line tool pdfinfo:

Fig. 02Reading the properties from the shell
pdfinfo contract.pdf
Title:          Service Agreement
Author:         Jane Doe
Creator:        Microsoft Word
Producer:       Adobe Acrobat 24.1
CreationDate:   Wed Aug 14 09:21:00 2026
ModDate:        Wed Aug 14 11:03:42 2026
Encrypted:      no
Page size:      595.276 x 841.89 pts (A4)
Metadata is not hidden in the usual sense — it is readable by any tool that knows the format. This is roughly what a recipient could learn in one command.Output structure of poppler's pdfinfo

Four ways to clean it

1. In the browser (no install)

ShredHub reads the PDF locally, clears the Info dictionary, removes the XMP stream and strips JavaScript triggers — then re-analyzes the result to prove it worked. Nothing is uploaded, so it also works for confidential documents.

2. Adobe Acrobat

Use Tools → Redact → Remove hidden information. It removes metadata, attachments and scripts in one dialog. Requires a paid plan for most versions.

3. Microsoft Word

If you still have the source document, File → Info → Check for Issues → Inspect Document removes personal data before you export. This is the only method that fixes the source, not just the export.

4. Command line (exiftool)

For batch jobs, exiftool -all= file.pdf strips most metadata. It needs installation and does not correctly handle every structure — verify the output afterwards.

Fig. 03Clean, then verify
  1. 01Read

    parse Info + XMP + objects

    ↓
  2. 02Clear

    empty the metadata fields

    ↓
  3. 03Re-scan

    confirm no fields remain

    ↓
  4. 04Share

    send the verified copy

Deleting metadata is a write operation; proving it is gone is a read operation. Do both, and keep the second result as evidence.

How to check your own PDF

  1. Drop the file into ShredHub — it stays on your device.
  2. Look at the Identity, Software and Rights groups.
  3. Clean with the standard preset and download the PDF.
  4. Re-scan it. A verified clean badge means the metadata is gone, not just hidden.

Takeaway

PDF metadata is easy to read, easy to remove, and very easy to forget. Add a cleaning step right before "send to external recipient" and the leak never happens.