Redact SSN-like numbers with PDF Redact
This opens a case file with two invented numbers and the SSN-like finder switched on. Press Find matches to see both with their pages, then mark them and check the preview.
- Open the PDF. The SSN-like pattern is already ticked.
- Press Find matches and read the list: each hit shows its page and context.
- Mark all or tick individual hits.
- Look at the page previews for numbers written another way, such as without dashes, and cover them with a box.
- Export, read the checks and download.
What to know
The pattern matches three digits, two digits and four digits separated by dashes or spaces, and skips shapes that cannot be real: an area of 000, 666 or 900 to 999, a group of 00, a serial of 0000. It is a format check only; it cannot tell whether a number belongs to a real person, and it will also match another number that happens to have the same shape. Numbers written without separators (123456789) are deliberately not matched because nine plain digits are very common; search for those with a regex, for example \b\d{9}\b, and look at every hit.
A full number is not the only risk. A last four digits next to a name can still identify someone, so decide what must go and mark it. The tool covers what you mark; it cannot decide what is sensitive.
Updated · Cosmovex
Questions
Does it check that the number is a real SSN?
No. It checks the shape and skips ranges that are not issued. It cannot know whether a number is real or belongs to anyone.
What about numbers without dashes?
They are not matched by the pattern. Use a regex such as \b\d{9}\b and review every hit, or draw a box.
What do I lose when the output is made of images?
Text in the whole output can no longer be selected, searched or copied, links and form fields stop working, and the file is usually larger. That is the cost of making sure the covered content is really gone, and the tool says so before you export.
How do I know the black boxes really removed the text?
After export the tool re-opens the file it just wrote and checks it: it contains only page images, there is no text layer to extract, and the pixels inside every box are black. You should still open the result and read every page; the checks cannot tell whether you covered everything sensitive.
The free plan covers everything on this page. PDF Redact Pro ($29, paid once) is described on the PDF Redact page.
