The fax should help with that, except that it'll reveal the location of your source fax. Using fax software/service needs its own security...
maybe snail-mail...
This is getting seriously nontrivial
If you do blow the whistel on the government, well, you better trust your journalist partner a lot. Or use a trustworthy lawyer as a go between.
But hey, one could send the fax back and forth, cut of the fax numbers and physically mail it from a very busy train station! And have a someone put the letter into the letterbox for you, someone you pay out of view of any CCTV cameras!
Nope. Microsoft Office has an option/plugin that lets you flag emails and documents as restricted. You can, for example, send such an email to someone else. That person can open the email and use a screenshot tool on that email. What they will get is a flat gray box where the email content was.
Not all pixels on the screen are the same. And you have no idea if your computer tattled on you and let the sending org know that you tried to screenshot the email.
This thread has nothing to do with preventing your source from finding out
This MS Office trick is a quick and easy way to show someone that you can’t trust Windows when it comes to screenshots. Ignore it at your own risk.
Right, I'd never use part of the original image. Best print it out and rescan it, preferably on different scanner.
Being truly anonymous is enormously difficult, even the Bauer pattern/Bayer filter shape in one's image sensor will give one away. That alone requires sophisticated filtering to decouple it from a specific sensor/camera.
And that's only the beginning, there are many other factors that contribute to the leaking one's identity.
1. First, most whistleblowing comes from B&W documents so color isn't an issue. If it is the case then there are alternative solutions which I'll not persue here in any detail (the post would be very long).
2. Printing then rescanning can decouple the camera to some extent as physical distortions are introduced by (a) the geometry of the inkjet nozzles or laser subsystem, (b) more geometric distortion is added by changes in paper size as its fused/dried, (c) additional distortion is added by the optics of the rescanning system, same goes for any further printing and so on. Also, each stage alters the gamma and adds other non linearities to the transfer.
3. Reducing the scanning resolution will add additional distortions, blurring is to be encouraged but not to the point of illegibility. (Doing a low resolution copy on a FAX machine with time/date/headers/footers turned off could help.)
4. This may sufficiently decouple the camera but it won't decouple the copying/scanning/rescanning equipment. Best do this on machines that are located well away from the source so they're not easily traceable or associated with you or your work.
5. Copying documents across multiple different photocopiers will help as they all add optical and other geometric distortions and transfer—gamma-type—errors. Deliberately lightening and darkening the image at each stage will further increase distortions. The combined distortions will make it very difficult (but not impossible) to reverse engineer the path back to the source. Have you ever wondered why the quality of the printed type in copies of evidence, FOIs, official documents etc. often looks so daggy and blurred as to be almost illegible?
6. The idea is to add as much distortion and blurring in the chain of copies as is possible and yet still keep the final copy legible/identifiable. It's a fine balance.
7. That still does not guarantee anonymity. Sensitive documents are often 'doctored' so that each recipient receives a subtlety different copy, spacing between words may change so may the page layout, also small seemingly insignificant and barely obvious 'marks' on the page will differ between copies thus identifying each recipient. None of the camera techniques above will overcome this problem.
8. One technique is to retype the document so that it is almost identical in both looks and content but with lots of subtle changes in both layout and wording—but NOT to the extent of changing the document's meaning. Same goes for modifying images (a complex matter). As a last resort this could give you a degree of plausible deniability in cases where marked documents are issued. If you do this it would be very advisable to spirit away actual copies of the originals in a truly safe location—you may really need them one day (after you're accused of, say, forgery). I'd suggest you only go down this path if you are very competent and the stakes are very high.
9. None of what I've said should be taken lightly. No matter how hard you try to decouple yourself totally from the source (or from any involvement) it is nigh on impossible to hide from formidable and sophisticated forensics if applied relentlessly.
10. Moreover, whistleblowers, more often than not, end up being reveilled by fellow workers and usually without a job or decent reemployment prospects. Unfortunately, our culture hasn't embraced whistleblowers and whistleblowing as it should. It's a very risky business. Take the case of whistleblower Daniel Ellsberg and the Pentagon Papers; his effort wasn't rewarded as we ought to expect and he nearly ended up being locked up.
Moreover, at one point Ellsberg wistfully noted that reaction by the American public to his efforts wasn't very favorable. When it comes to whistleblowing it seems not much has improved in the last five decades.
_
Edit: FYI, there are many excellent videos on Daniel Ellsberg and whistleblowing on YouTube.
But if I ever learned anything about that area it's that there's probably still five other details I haven't considered.
I think this is a little unfair. I would not blame anyone for thinking that the part of the image they cropped is deleted. That's what the word means, that's what it looks like is happening, there's usually no warning that it's recoverable.
When you delete text it is really gone from the document, even though you can undo to get it back. It's totally reasonable to expect images to work the same way.
Those are things many recent UI "improvements" haven't considered.
Abusing Microsoft Office applications to do things they're not designed to do has sort of a bathtub curve of likelihood versus tech literacy. The people most likely to do it are either largely tech illiterate or skilled hackers with far too much time on their hands.
I’m well aware that it’s reversible so I’m careful of what I put in to begin with, but I’m not going to open an image editor unless I really have to
I have other shit to do
Kevin in accounting might be making a document for work and might take a screenshot of his entire desktop to put in an onboarding document, intending to crop it. He could have a password in Notepad off to the side
My significant other is outside of tech and wasn't aware of either of these. Once shown, there was no reason to go back to using anything else to crop images.
No image editors involved!
saving a screenshot as JPG (lossy format) adds more loss - if you started with a lossy image in the first place then we are losing even more information by screenshotting it and saving it again in a lossy format.
even when not intended for print, when making formal documents and presentations I look for the original images at highest resolution available.
- Bluetooth file transfer… I don’t even know where to begin
- AirDrop requires you to be in the Apple ecosystem. I think default settings also require you to be logged in the same Apple account on both devices too
- Cloud storage services need you to install the app or login into the website on both your phone and the computer with the same account. Very few of my non-tech friends even have any cloud storage service outside what their phone uses
Ngl, I take photos of my screens sometimes
whatsappweb opened in a browser
pushbullet
google photos / Gmail to yourself
For example, "motion photo" feature I used few years ago would give you perfectly normal JPG file that also happened to store a short video. Nothing I know other than Samsung phones can read the video part directly, though there exists desktop software that can extract it and save as separate video file.
More recently, I've noticed photo editor tools on Samsung phones allow you to undo previously made edits on an already saved file, including any destructive crops and transforms - meaning the original is being persisted somewhere. I haven't checked where, but I imagine it's again attached to the back of the JPG you thought you've cropped/redacted.
Somewhat related: PlantUML. It's a tool for generating diagrams from plaintext descriptions. One of its most interesting, if not well-known, features is that it saves "editable PNGs" by default - that is, PNGs it produces have the plaintext diagram code attached to the image as metadata. Seems mostly harmless, but it could technically leak information you put in the diagram code but made not visible in the image.