How to remove the creation date from a pdf's metadata
- Step 1Read the current dates first — Open the PDF in Acrobat (File → Properties → Description) or run
exiftool file.pdfand note the Created / Modified values. This is the timeline you are about to erase. - Step 2Drop the PDF onto the scrubber — Add the file in the tool above. It loads locally with pdf-lib — bytes are read into memory in your browser, never uploaded.
- Step 3Let it run automatically — The scrubber is a no-options tool, so it runs the moment the file is added. There is no field where you choose 'dates only' — the date fields are reset and the other six metadata fields are cleared together.
- Step 4Download the scrubbed PDF — Save the cleaned copy with the download button. The original on disk keeps its real dates; the download has both timestamps reset to epoch.
- Step 5Verify the dates read 1970-01-01 — Re-open the file: Acrobat's Created and Modified should both show 01/01/1970.
exiftoolwill printCreateDate: 1970:01:01 00:00:00Z. That epoch value is the intended result, not a bug. - Step 6Check for an XMP date if your viewer is strict — If a viewer still shows a real date, it is reading
xmp:CreateDatefrom the XMP packet, which this tool does not change. Re-save through pdf-compress-lossless to rebuild the document and drop the stale XMP dates.
How each date / metadata field is treated
The scrubber processes all document-information fields in one pass. The two date fields are reset to the Unix epoch; the rest are emptied.
| Field | Before (example) | After scrubbing |
|---|---|---|
/CreationDate | D:20260214094205Z (2026-02-14) | Reset to 1970-01-01T00:00:00Z |
/ModDate | D:20260601173000Z (2026-06-01) | Reset to 1970-01-01T00:00:00Z |
/Author | J. Drafter | Empty string |
/Creator | Microsoft Word | Empty string |
/Producer | Adobe PDF Library 24.1 | Empty string |
/Title | Memo-final-v7 | Empty string |
/Subject | Internal use only | Empty string |
/Keywords | draft, q1, confidential | Cleared |
Date fields and whether this tool reaches them
A PDF can express a date in several places. This tool resets the two document-info date entries only.
| Date location | Reset by this tool? | Note / alternative |
|---|---|---|
Document-info /CreationDate | Yes — set to epoch | Shown as 'Created' in Acrobat |
Document-info /ModDate | Yes — set to epoch | Shown as 'Modified' in Acrobat |
XMP xmp:CreateDate / xmp:ModifyDate | No — XMP not rewritten | Re-save via pdf-compress-lossless to drop XMP |
XMP xmp:MetadataDate | No | Same — lives in the XMP packet |
| Annotation timestamps (comment dates) | No | Remove with pdf-annotation-remover |
| A date printed in the visible page text | No — that's content, not metadata | Redact with pdf-pii-redactor |
Cookbook
Real timestamp before/after snapshots as read in Acrobat and ExifTool. The epoch value (1970-01-01) is what the tool intentionally writes.
A 'recent' memo with an old Created date
A memo claimed to be written this week, but its /CreationDate is three months old because it was drafted from a saved template. Resetting the date removes the contradiction.
Before (Acrobat → Description): Created: 2026-03-02 08:11 Modified: 2026-06-05 14:50 After scrubbing: Created: 1970-01-01 00:00 UTC Modified: 1970-01-01 00:00 UTC
FOI release where ModDate reveals a late edit
A document disclosed under FOI had a modification stamp hours before release — evidence of last-minute redaction tweaks. Scrubbing resets ModDate so the edit window is no longer inferable from metadata.
Before: Created: 2025-11-20 10:00 Modified: 2026-06-06 23:58 <- night before release After: Created: 1970-01-01 00:00 UTC Modified: 1970-01-01 00:00 UTC
ExifTool shows epoch after the scrub
Confirming with ExifTool: the document-info dates collapse to epoch. If an XMP CreateDate is still real, you will see it on a separate line and know to re-save losslessly.
$ exiftool scrubbed.pdf | grep -i date Create Date: 1970:01:01 00:00:00Z Modify Date: 1970:01:01 00:00:00Z # If XMP persists you might also see: # Metadata Date (XMP): 2026:06:06 23:58:01 # -> re-save via pdf-compress-lossless
Why not a blank date? (it's reset, not removed)
Some users expect an empty Created field. pdf-lib's API sets a Date value, and the tool uses new Date(0) — the epoch — rather than deleting the key. The practical effect is the same: the real date is gone.
Implementation: doc.setCreationDate(new Date(0)) // 1970-01-01T00:00:00Z doc.setModificationDate(new Date(0)) Viewer shows: Created: 01/01/1970 (not blank) The original timeline is unrecoverable from the info dict.
Date is printed on the page, not in metadata
A letter has '14 February 2026' typed in the body. Scrubbing resets the metadata dates but the visible date stays — it is content. Redact it instead.
Metadata Created: 1970-01-01 (scrubbed) ✓ Letter body: 'Dated: 14 February 2026' ✗ still visible Fix: /pdf-tools/pdf-pii-redactor to black out the printed date
Edge cases and what actually happens
Dates show 1970-01-01 instead of blank
ExpectedThe tool resets both date fields to the Unix epoch (new Date(0)), so viewers display 1970-01-01 rather than an empty value. This is intentional — the real creation/modification timeline is unrecoverable from the document-info dictionary. If a strictly-empty date is required, you would need a dedicated PDF editor that deletes the keys outright.
XMP CreateDate still shows a real date
XMP not rewrittenThis tool resets the document-info dates only; it does not touch the XMP metadata packet's xmp:CreateDate / xmp:ModifyDate. A few viewers prefer XMP. If a real date persists, re-save through pdf-compress-lossless to rebuild the file without the stale XMP.
Comment / annotation timestamps remain
Out of scopeEach annotation can carry its own modification date. Those live in the annotation objects, not the document-info dictionary, so they survive a metadata scrub. Strip them with pdf-annotation-remover.
A date is printed in the visible page content
Not metadataA date typed into a letter, invoice, or signature block is page content, not metadata. The scrubber will not change it. Use pdf-pii-redactor to redact a visible date.
Digitally signed PDF — signing time visible
Signature breaksA cryptographic signature embeds its own signing time and is invalidated by any re-save. Scrubbing the metadata will break the signature without removing the embedded signing time inside it. Inspect with pdf-signature-verify; re-sign after scrubbing if the document must stay signed.
File exceeds the free 2 MB limit
BlockedFree handles files up to 2 MB; larger files are blocked before processing with an upgrade prompt. Pro raises the ceiling to 50 MB. The date-reset itself is cheap, but the file still has to load fully into the browser.
More than 50 pages on free tier
BlockedThe free page ceiling is 50 pages (Pro: 500). The runner checks the page count on load and blocks before scrubbing if you exceed your tier.
PDF had no date metadata to begin with
ExpectedIf the source carried no Created/Modified values, the tool simply writes the epoch into both fields. The output is a valid PDF whose dates carry no real information either way.
You wanted to set a specific custom date
Not supportedSetting an arbitrary date (e.g. backdating to a chosen day) is not supported — the tool only resets to epoch. Use a desktop PDF editor or a scripting library directly if you must write a chosen timestamp.
Frequently asked questions
Does the tool delete the dates or reset them?
It resets them. Both /CreationDate and /ModDate are set to new Date(0) — the Unix epoch, 1970-01-01T00:00:00Z. Viewers will show 01/01/1970 rather than an empty field. The practical effect is the same: the original drafting timeline cannot be recovered from the document-info dictionary.
Why does my viewer still show a real creation date after scrubbing?
It is reading the XMP metadata packet (xmp:CreateDate), which this tool does not rewrite. Re-save the file through pdf-compress-lossless to rebuild it without the stale XMP dates, then re-check.
Can I set a custom date instead of removing it?
No. This tool only resets both date fields to the Unix epoch — there is no date picker. To write a specific timestamp you would need a desktop PDF editor or a scripting library that lets you set the value directly.
Will removing dates affect a digital signature?
Yes. Any re-save changes the file bytes and invalidates an existing signature, and the embedded signing time inside the signature is not removed by scrubbing. Reset dates before signing, or verify and re-sign with pdf-signature-verify afterwards.
Are there other date fields I should worry about?
Yes — the XMP packet (xmp:CreateDate, xmp:ModifyDate, xmp:MetadataDate) and per-annotation timestamps. This tool reaches neither. For XMP, re-save losslessly; for annotation dates, use pdf-annotation-remover.
Does it also remove the author and software at the same time?
Yes. The scrub is a single pass that also empties Author, Creator, Producer, Title, Subject, and Keywords. See the remove-author and remove-software-info guides for those angles.
Is the document uploaded anywhere?
No. The reset runs in your browser with pdf-lib. The file and its timestamps never leave your device; only an anonymous run counter is recorded for signed-in users.
Will the visible page change?
No. Dates are metadata, invisible to readers. Only the hidden Created/Modified stamps and the other document-info fields change; the pages are identical.
A date is printed in the body of my document — will this hide it?
No. A date you can see on the page is content, not metadata. Redact it with pdf-pii-redactor; the metadata scrubber only touches the hidden timestamp fields.
What's the largest file I can scrub?
Free: 2 MB and 50 pages. Pro: 50 MB and 500 pages. Pro+Media: 500 MB and 2,000 pages. Because the operation only rewrites metadata, runtime is dominated by loading the file, not by page count.
Can I scrub an encrypted PDF's dates?
A permissions-only encrypted PDF loads and scrubs (the tool opens files with encryption ignored). A PDF with an open password must be unlocked first via pdf-remove-password.
Can I automate this for a batch of files?
Yes. Pair the @jadapps/runner and POST each file to 127.0.0.1:9789/v1/tools/pdf-metadata-scrubber/run. The tool's option schema (from GET /api/v1/tools/pdf-metadata-scrubber) is empty, so the payload is just the file. Processing happens locally.
Privacy first
All PDF processing runs locally in your browser using PDF-lib and pdf.js. No file is ever uploaded — only metadata counters are saved for signed-in dashboard stats.