Redact PDF (Permanently Remove Sensitive Information)
Black out text and areas for good — and strip the hidden data that gives secrets away.
Boxes
Find and mark
Never stored: they are forgotten when you leave the page.
Your own pattern
Your redacted PDF is ready
Also removed from the file:
For general information only, not legal advice. Templates are generic starting points — have a qualified lawyer review anything you rely on.
About the Redact PDF (Permanently Remove Sensitive Information)
Remove information from a PDF so that it is really gone. Draw boxes over anything on the page, or let the tool find email addresses, phone numbers, card numbers (Luhn check), IBANs (check digits), PAN and Aadhaar numbers, US Social Security numbers, your own words and names, or your own pattern — then review every find before redacting. A US court filing option follows FRCP 5.2 and keeps only the last four digits of SSNs and account numbers and the year of birth dates.
A black rectangle drawn over text is not a redaction: the text stays in the file and can be copied or uncovered. So every page with a box is redrawn as an image with the boxes burned in — its text, pictures and drawings are destroyed, while pages without boxes stay as they were. Hidden information is removed from the whole file: document information and metadata, comments, attachments, form data, JavaScript, hidden layers and more. Afterwards a check searches the result for everything you redacted. Your file never leaves your browser.
How to use it
- Drop a PDF onto the box above, or tap Choose a PDF to redact.
- Under Find and mark, tick the kinds of data to look for, add words and names (one per line) or your own pattern, and press Find. Review the list: untick anything that should stay.
- Draw boxes over anything else — signatures, photos, stamps, handwriting — with Draw, or use Redact this whole page. Select lets you move, resize or delete boxes (Delete key). With the keyboard, focus the page and press Enter to add a box, then move it with the arrow keys and resize it with Alt and the arrow keys.
- Choose the box colour and the resolution of redacted pages; tick Keep redacted pages searchable to recognise their remaining text again (OCR).
- Press Redact PDF. Read the check, then Download and open the file to look it over before you share it.
Examples
Find: card numbers, IBANs, phone numbers · words: your name and address lines
Those pages become images with black boxes; the account details, metadata and the bank’s hidden form data are gone from the file
US court filing option · SSN 123-45-6789, Account No: 1234567890, Date of Birth: 12/03/1985
███████6789, ██████7890, ██████1985 — the last four digits and the year stay visible, as the rule allows
Find: Aadhaar numbers, PAN · draw a box over the photo and the QR code
Only numbers that pass the Aadhaar check digit are found; the photo and QR code are destroyed with the boxes
Common uses
- Sending ID documents, statements or salary slips with the numbers that matter blacked out.
- Publishing reports, case files and RTI or FOI responses without personal data.
- Preparing court filings that may show only the last four digits of identification and account numbers.
- Removing names and contact details from CVs, research data or customer documents before sharing them.
Why redacted pages become images
Text in a PDF is stored as characters, so anything that only covers it — a black rectangle, a highlight, white paint — leaves it in the file, where it can be copied, searched or uncovered by moving the cover. This is a common redaction mistake, and the reason guidance such as the NSA’s Redacting with Confidence (report I333-015R-2005) insists that the underlying content be removed.
Here, each page with a box is drawn in your browser at the resolution you choose (150–300 dpi), the boxes are painted into the pixels, and the page is replaced by that picture. Nothing of the old page remains: no text, no fonts, no vector drawings, no images, nothing outside its visible area. Pages without boxes keep their real text and quality.
What is removed from the whole file
- Document information and XMP metadata — title, author, subject, keywords, software and dates.
- Comments and annotations — notes, highlights, stamps, drawings, file attachments and redaction marks; tick Keep the look of comments to draw them into the pages instead (their text is still removed).
- Form data — fields are flattened: what they show stays as page content, the form and its stored values go.
- Attachments, JavaScript and automatic actions, links that run scripts or point to a redacted address, and redacted words in bookmark titles (replaced by █).
- Hidden layers, page thumbnails (small pictures of the old pages) and the tag structure, whose alternate texts can repeat the content.
- Unused resources: pages often share one list of fonts, pictures and drawings, so a picture or text block of a redacted page would stay in the file through the other pages. Every page that stays keeps only what it actually shows.
Redaction marks made in another app but never applied are turned into boxes when you open the file, ready to burn in.
How finding works
The text of every page is read with the exact position of each character — including text in form fields, comments, hidden layers and the invisible text layer of scanned pages that went through OCR — so a box covers precisely the characters found, a little padded. Numbers are only reported when they pass their check: card numbers the Luhn check digit (ISO/IEC 7812-1), IBANs their MOD 97-10 check digits (ISO 13616), Aadhaar numbers the Verhoeff check digit; US Social Security numbers must be written 123-45-6789 (or follow “SSN”) and not use the area, group or serial numbers the SSA never assigns (000, 666 or 9xx; 00; 0000 — SSA, POMS RM 10201.035). Card numbers are found even when they are printed right next to other digits, such as an expiry date, a CVV or a row number. Phone numbers are recognised with libphonenumber, for the country you choose. Scanned pages without a text layer can’t be searched: draw boxes there, or run PDF OCR first.
US court filings: FRCP 5.2
Federal Rule of Civil Procedure 5.2(a) says a filing that contains a Social Security number or taxpayer-identification number, a birth date, the name of a known minor or a financial-account number may include only the last four digits of the identification number, the year of birth, the minor’s initials, and the last four digits of the account number (Cornell LII, Rule 5.2). The court-filing option covers SSNs, card numbers, IBANs, numbers written after “Account”, “Acct” or “A/C”, and dates written after “Date of birth”, “DOB” or “Born”, leaving the last four digits or the year visible. Minors’ names can’t be recognised automatically: add them under Words and names (the whole name is blacked out). The rule has exemptions (Rule 5.2(b)) and courts can order otherwise — this tool doesn’t decide what your filing needs.
The check after redacting
When the file is ready it is opened again and searched for every item you redacted, and the detectors run once more: on the text of the pages that stayed, on links, bookmarks and document information — and, for the redacted pages, on what their pictures still show (the text no box covers). So a box that was moved off a finding, made too small or deleted is reported, as is anything the detectors find that you didn’t tick. The file’s structure is checked too, for leftover metadata, scripts, attachments, form data and comments. Items you unticked on purpose are reported as such. The check can only look for what was found or typed, and it can’t read text that is part of a picture — look over the result yourself before sharing it.
Limitations
- Redacted pages become images: their text can’t be selected (unless OCR is on), and the file usually gets larger.
- Text that is part of a picture (scans, photos, screenshots) is found only through an OCR text layer; otherwise draw boxes over it.
- Detectors can miss information written in unusual ways, and can’t know everything that is sensitive in your document; always review the result.
- Removing the tag structure makes the redacted file less accessible to screen readers.
- Password-protected or restricted PDFs must be unlocked first with Unlock PDF.
- Redacting changes the file, so digital signatures will no longer verify.
- This tool is a technical aid, not legal advice: what must be redacted depends on the law, the court or the organisation you send the document to.
Privacy
Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot. Only if you tick “Keep redacted pages searchable”: the text recognition engine and the language data (a few MB) are downloaded once from this site. Your file is never uploaded.
Frequently asked questions
Is the redaction permanent?
Yes. Pages with boxes are replaced by pictures with the boxes painted into the pixels, so the covered text no longer exists in the file — it can’t be copied, searched or uncovered.
Why not just draw black boxes in a PDF editor?
A drawn box only covers the text; the text stays in the file and anyone can copy it or delete the box. Use PDF Editor white-out for corrections, and this tool for anything confidential.
Does it find Aadhaar and PAN numbers?
Yes: PAN numbers by their format (five letters, four digits, one letter) and Aadhaar numbers (12 digits, with or without spaces) only when their Verhoeff check digit is valid, so ordinary 12-digit numbers are not blacked out by mistake.
Can I redact a scanned PDF?
Yes, by drawing boxes. To find text in a scan, it needs an OCR text layer: run PDF OCR first, then search here — the boxes then cover the scanned words.
Are my words and the found data stored anywhere?
No. The PDF, your words, your pattern and the results stay in this browser tab and are forgotten when you leave the page; only settings such as the box colour and the kinds of data to look for are remembered.