Hacker News Reader

Comment by unimpressive | Hacker News Reader

I've wondered this for a while, and a web search was not forthcoming, so I'll ask here:

How do you properly redact a physical paper document? I would imagine that black sharpie doesn't cut it.

Where I work, the physical document is first scanned into an electronic form (multipage formats like TIFF or PDF, usually). The electronic document is then imported into a proprietary redaction program (there's quite a few on the market, most copier/scanner vendors have an offering).

It then assists with the initial redaction by performing OCR on the content, allowing you to quickly search for sensitive information. The software provides built-in tools for blacking out sensitive sections of the document - some even provide configurable overlays/stamps on the blackouts for why that section was redacted.

Then one or more humans manually go over the documents to make sure that nothing was missed. When they are happy with the results, the document is finalized into a TIFF or PDF document with the pages flattened into single image objects (so you can't just delete the black boxes - they are now part of the embedded image).

The original documents do not need to be harmed in any way, they just need to be able to be scanned to start the process.

Conversion to an image format helps to ensure that if the source was an electronic document that sensitive original source metadata or proprietary document format binary blobs don't inadvertently leak redacted information. If a physical copy is required, you just print it.

>When they are happy with the results, the document is finalized into a TIFF or PDF document with the pages flattened into single image objects (so you can't just delete the black boxes - they are now part of the embedded image).

And if you fail to do that, you get this (original link is gone, but here's the discussion, and the gist is the TSA released "redacted documents" where you could just select over the blacked out areas and ctrl-C to recover the text).

https://news.ycombinator.com/item?id=980465

The UK National Archives suggest a few different methods on pp. 13-14 of this PDF: http://www.nationalarchives.gov.uk/documents/information-man...

1. I immediately read most of this document

2. Appendix 1 (the section referenced) might be objectionable to the eco-conscious. Tread with caution.

First, you █████████ ███ █████. Then, after █████ ████ █████ ████ the █████, you can either ████ █████ █████ ████ ██ ███████ ███, or ████ █████ ██████ ████ █ ███'█ ████ ██.

With a razor blade and a photocopier.

So you cut out the parts you don't want seen and then put a sheet of black underneath?

Or no sheet underneath at all.

At one point some documents were "redacted" and released. Turned out you could just remove the black boxes in the PDF and see the text underneath.

I would guess that it's not worth the effort, due to the risk of a physical process failing (lose painter's tape, need to make a copy for cutting, etc).

In modern practice, scan, bitmap paint 100% black, print?

In the very few documents I've had to redact (PII, not tyranny :P) I actually used white boxes instead of black to waste less toner, but yeah it's the same idea.

Black sharpie then photocopy?

Edit: actually, just put a black filter over the sensitive parts while photocopying. That way you don't damage the original and you can send the photocopy over.

You also have to mess with the contrast levels or set it to "black and white" instead of grayscale. The default grayscale settings on photocopiers tend to leave the copies still somewhat legible (I probably wasn't using the "proper" sort of marker, one of those heavy duty aluminum-can permanent ones.)

I think the redactions are actually done using a narrow painter's tape, then a photocopy of the originals. The painter's tape won't stick to the paper, but shows up as black on the photocopy.

Reply on news.ycombinator.com