How-to Guides Forensics

How to Check a PDF for Hidden Changes

Find out whether a PDF was edited after it was created or signed: incremental saves, changed text between revisions, weak redactions and signature coverage.

Last reviewed:

AdSense Placeholder
Slot: header_reference_page
On this page

Steps

  1. Hash the file you received

    Record the SHA-256 of the PDF with the File Checksum Calculator. If you still have the copy you sent out, hash it too: identical checksums mean not a single byte changed and you can stop here.

  2. Open it in Document Forensics

    Drop the PDF into Document Forensics. The Overview shows the number of revisions, the producer and a list of what stands out, such as "Edited after creation".

  3. Walk through the revision history

    Open Revision History. Every incremental save is listed oldest first, with its byte range, producer, modification date and how many objects it added, changed or deleted.

  4. See what text changed between saves

    Open What Changed, pick two revisions in Compare from and To, and read the Text added and Text removed for each page.

  5. Run the redaction check

    Open Redaction Check. It lists text that is still in the file underneath a box painted over it, and redaction marks that were added but never applied.

  6. Check what each signature covers

    Open Signatures. Covers the whole file means nothing was appended after signing; Covers only the first ... of ... means later bytes, and possibly later edits, are outside the signature.

  7. Read the metadata

    Open Metadata to see the information dictionary and XMP packet: creator and producer software, creation and modification dates. A modification date earlier than the creation date is flagged.

  8. Compare against your own version

    If you have the original document, copy the text of both versions into the Text Diff Checker and press Compare to see every added and removed line.

What you need

The PDF exactly as you received it, not a printed or re-exported copy, and ideally the version you sent. The analysis runs in your browser and the document is never uploaded. The same tool also reads Word, Excel and PowerPoint files, where it shows tracked changes, comments and hidden text.

Reading the result

A PDF can be changed in two ways. Most editors that rewrite the whole file leave no history behind. But many tools, including those that add comments, fill forms or sign, use an incremental save: they append the changed objects to the end of the file and leave the original bytes in place. Document Forensics reconstructs each of those saves.

Here is a typical pattern for a signed agreement that was quietly modified:

Tab What it shows
Overview 3 revisions; more than one producer wrote the file
Revision History Revision 1: original save. Revision 2: the signature. Revision 3: a different producer, a few objects changed
What Changed (2 → 3) Page 2: Text removed payment within 30 days, Text added payment within 90 days
Signatures Covers only the first part of the file

Read it in order. Revision 2 is where the signature was applied, so the signature covers revisions 1 and 2. Revision 3 came later, from different software, and changed the payment terms on page 2. Many PDF readers show the latest version without any warning, so the reader sees "90 days" next to a signature that was made over "30 days".

A file with a single revision and a signature that covers the whole file is the clean case. A file with one revision and no signature tells you little: it may have been rewritten completely by an editor that keeps no history.

Common pitfalls

  • Calling every extra revision tampering. Signing itself adds a revision, and so do form fills and annotations. Look at what changed, not how many saves there are.
  • Assuming no history means no edits. "Save As" in many editors, printing to PDF and online converters produce a fresh file with one revision. Only a comparison with your own copy, or a checksum, rules out changes.
  • Over-reading the redaction check. Table shading and figure overlays also sit on top of text. Each result is a lead to look at, not a verdict.
  • Treating "signature present" as "signature valid". This tool reports coverage only. Verify the certificate and signature in a validating reader.
  • Analysing an encrypted file. The structure of an encrypted PDF is readable, but its text and streams are not, unless it uses an empty user password. The Overview says which case applies.

Going further

For contracts and invoices, keep the SHA-256 of every version you send. A later dispute then comes down to comparing two hashes, and Document Forensics shows exactly what differs when they do not match.

Try these tools

See also

  • Glossary PDF (Portable Document Format)
    PDF is a file format for documents that look the same on every device, bundling text, fonts, images and layout into one file.
  • Glossary Digital signature
    A digital signature is a value computed from a message and a private key that proves who created it and that it has not been changed.
  • Glossary Metadata
    Metadata is data about data: the hidden or secondary information a file carries about itself, such as its author, creation date.
  • Glossary Hash
    A hash is the fixed-length fingerprint a hash function computes from any input.

Frequently Asked Questions

Not necessarily. Filling in a form, adding a comment, signing or simply saving again in some editors all append a revision. What matters is what each later revision changed, and whether it came after a signature.

No. It shows which signature fields exist and how much of the file each one covers. Checking the certificate and the cryptography needs a signature validator, such as the one built into a PDF reader.

AdSense Placeholder
Slot: footer_leaderboard