-3

Apologies in advance if this topic is too vague, I wanted to get an idea of peoples' opinions.

Technology now exists that makes photography a unreliable medium in terms of truth.

Image and video editing software allows people with a small amount of understanding to create false images that may mislead non-computer literate people to "believe their own eyes". This can lead to mis-information, to slander, and to propoganda even.

Is it the responsibility of photo editing software companies like adobe to label edited photos as such ? Or should a company like Google / Mozilla / facebook use meta-data / photo analysis on their platforms, or even AI to ensure that users know whether what they are seeing is real, or fake ?

closed as primarily opinion-based by Steffen Ullrich, Anders, CaffeineAddiction, forest, Jeroen - IT Nerdbox Apr 17 '18 at 4:54

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • I believe it is not possible to give you a correct answer considering all possible scenarios over here. Each service that you use has terms of service and if they talk about authenticity of the data in the terms of service, then they are on the hook for ensuring it otherwise they are not responsible. If they do, good for them, otherwise well, they are not responsible – Limit Apr 15 '18 at 15:31
  • The burden of proof is on the one making the claim. So if a newspaper publishes an image and claims it is real and it is found that it is not. It is the newspaper that has to either admit fault or pass the blame. – Dan D. Apr 15 '18 at 15:48
  • 4
    I'm voting to close this question as off-topic because this is not a technical question. And it is not a new question either: falsifying evidence or news by editing photos, creating false videos etc is already done for a long long time and the new thing is only that it got much easier. The question is even if somebody is responsible to mark fakes as such in the first place (and what are the legal and enforcable consequences if not done) or if one should simply learn what can be faked and learn to distrust possible fakeable evidence. – Steffen Ullrich Apr 15 '18 at 15:51
  • 3
  • It could be an IT security issue if you took the perspective that false images could be used in phishing campaigns. Then the question is, how to detect them and notify the user that what they're looking at isn't real... preferably without breaching Confidentiality or Integrity. – nbering Apr 15 '18 at 16:33
4

Photography has never been a reliable medium for truth.
The techniques to manipulate images are almost as old as photography itself. Most movie effects have been done with analog green-screen techniques, so even video can be made to lie with relatively primitive means.

Even if all commercial software watermarked its images, there's FOSS. Camera sensor signatures can be analyzed, but they can be simulated as well. Cryptographic camera signatures could provide verification for some images, but the public will eat up sensational images signed or unsigned. And against all digital protections, there's the analog loophole.

In information security - since that's the community you're asking - the guiding principle is "trust no one", embodied in the concept of zero-trust security. Treat every image you see as just a drawing, unless independently verified and corroborated by out-of-band evidence.

1

Is it the responsibility of photo editing software companies like adobe to label edited photos as such?

Technology always gets better. It never gets worse. What's today within the reach of Adobe to do, is within the reach of hobbyists to do tomorrow. If Adobe enforces labeling, it may delay the problem for some small amount of time, until Open Source software are made with the same capabilities.

Or should a company like Google / Mozilla / facebook use meta-data / photo analysis on their platforms, or even AI to ensure that users know whether what they are seeing is real, or fake?

This sounds more like science fiction than anything else. Is there meta-data that cannot be faked, like the photo itself? It's far easier to fake the string 'Canon EOS 1D' than to fake the photo... Often it's easier to find third party verification or debunking than proving if a photo is edited or not.

  • 1
    I have seen demonstrations of software that can identify patterns in photographs that are like fingerprints left by various image manipulation algorithms. The software was able to correctly identify regions that were stretched or touched up to blend two photos together. So it’s not impossible to forensically identify fakes, but that kind of detection would just result in an arms race of detection and obfuscation techniques. – nbering Apr 16 '18 at 12:10

Not the answer you're looking for? Browse other questions tagged or ask your own question.