PDF metadata: what it reveals and how to review it

Black-box redaction does not clear Document Properties, and Merge PDF or Compress PDF does not reliably strip them either. After export, Author can still show an account name or an email, including a login left behind after the visible names were covered. Open File, then Properties, on the file you are about to send. If the document is leaving the organization, sanitize those fields on purpose. Split PDF does not do that step for you.

People miss metadata until someone opens Properties. It is the same class of leak as a phone number written on an envelope that looked blank.

What it is

Metadata describes the document. It is not the body text on page 1.

| Field | Example | Risk | | --- | --- | --- | | Title | Draft Contract v4 | Shows that the file is still a draft | | Author | jane@firm.com | A personal or internal ID | | Subject | Project Falcon | A codename | | Creator | Microsoft Word | Low | | Producer | Adobe PDF Library | Low | | Created | 2026-03-14 | A timeline | | Modified | 2026-03-18 | Shows a recent edit |

Some files also carry XMP: camera EXIF from an embedded photo, custom XML from an enterprise tool, and sometimes a location from a phone scan.

Printing to PDF again does not guarantee a clean file. The new export may keep the old fields or add another software name on top.

When it matters

• Legal and M&A drafts shared outside the firm
• Anonymous peer review
• A handoff to a journalist or a whistleblower
• HR documents leaving the organization
• An RFP response with an internal codename in properties
• A compliance packet whose audit trail is useful inside and harmful outside

Inside a team, metadata helps version tracking. The leak is on the way out.

How to look before you send

Desktop readers

Adobe Acrobat or Reader: File, then Properties, then Description, Initial View, and Advanced.

Foxit, PDF-XChange, and Preview on Mac put this in an Inspector or Get Info panel. Look for "More Info" or "Document Properties."

Write down every field. If Author is a Windows username like DESKTOP-JSMITH, fix it before you send.

Merge and compress

Merge PDF may keep the author from the first source file. Compress PDF rarely sanitizes.

Do it in this order: sanitize, then merge, then compress, then send.

Uploads

A random converter may log or keep metadata on a server even if the download looks clean. Treat that like the risk in browser PDF privacy.

Sanitize, then check again

1. Export with clean properties. Set Author and Title in Word or Google Docs when the export settings allow it.
2. Open Properties and write down what is there.
3. Sanitize with a tool that removes metadata (Acrobat Pro "Remove Hidden Information", some command-line tools, or a privacy scrubber).
4. Open Properties again. Author should be blank or a generic organization name.
5. Redact the page separately. Clearing metadata does not redact content. Try to copy text from under any black box.
6. Add a password if you need one. Encryption does not hide metadata equally in every tool, so sanitizing still matters (password guide).
7. Rename the file. Smith_Jane_Review.pdf leaks on its own.

Metadata, encryption, and redaction

| Control | Hides page content | Hides properties | Stops forwarding | | --- | --- | --- | --- | | Metadata scrub | No | Often yes | No | | Redaction | If it is done correctly | No | No | | Password | When the file is locked | Partly | No | | Portal permissions | Sometimes | Sometimes | Better |

Use the combination the risk actually needs. A password alone does not cover properties or page text.

Mistakes

"Save as PDF" from a browser is not automatically clean. The print step adds its own producer string, and the author can still come from the HTML meta tags.

A black highlight in a Word export is not redaction in the PDF.

A merged file can inherit the wrong author. In some tools the first file wins. Scrub the sources, or put them in an order you have checked.

A PDF can contain embedded files, and those files have their own metadata.

A firm checklist

Some firms scrub metadata on every outbound PDF, as part of ISO 27001 or a client DPA. A practical list:

• [ ] Author is generic or blank
• [ ] Title does not reveal the matter
• [ ] Custom XMP has been reviewed
• [ ] Embedded files are gone
• [ ] The filename is neutral

Related reading

• PDF Password Protection
• How Browser PDF Tools Protect Privacy
• Combining Scanned and Digital PDFs

Limits

Clearing Document Properties does not redact visible or hidden page text. If you are removing names or account numbers from the pages, test by selecting the text.

An embedded image can still carry EXIF (camera model, GPS) after the PDF Title and Author fields are cleared.

Sources

• Adobe, PDF properties and metadata: common document information fields
• ISO 32000 overview (PDF Association): PDF as an ISO standard
• How Browser PDF Tools Protect Privacy: client-side processing and upload risk

Last reviewed: September 2026. Sources and scope checked for accuracy.

Review author, title, and dates before a sensitive send. Sanitize on purpose. A black box and a password do different jobs, and the filename is metadata too. Clean pages are not enough if Properties still names the account.