2

Hosting copyrighted material without permission is illegal in most countries. However, what happens if two independent websites host seemingly random data, which combined (by XOR-ing) form a stream of copyrighted material?

The idea is rather simple, someone has digital copyrighted material, the digital form might be 011001 (CR). He then creates a random string of binary digits, for example 010111 (XOR1) and XORs the copyrighted material with the original, resulting in 001110 (XOR2). Both XOR1 and XOR2 seem random and unrelated to CR, but when XOR-ed, they form CR.

Now this person anonymously uploads XOR1 and XOR2 to independent hosts, and publishes links to both on a third independent forum.

I would assume the anonymous person violates the copyright, but do any of the three hosts do so as well? Can the copyright holder force the removal of XOR1 or XOR2, since for neither XOR1 nor XOR2 it is provable that it is derived from CR and independently both look random.

Moreover, is the post on the forum illegal, as it describes how to get the copyrighted material, but doesn't host it itself?

I am curious about this w.r.t. either Dutch or US law.

Improve this question
1

1 Answer 1

Reset to default
3

It infringes the copyright. It can easily be proved that both XOR1 and XOR2 derive from the source work by XOR-ing the streams with each other. It's just like any encrypted copy: it infringes the copyright, but only those who can decrypt it are in a position to know that it infringes the copyright.

The posts on the forum are illegal because they infringe the copyright; it doesn't matter that they are derived works rather than the work itself, just as your drawing of a copyright-protected image infringes copyright because it is a derived work without being the work itself.

Improve this answer
10
  • Not a lawyer so I won't up-vote but I certainly hope this answer is correct.
    – Andrew Steitz
    Commented Sep 9, 2017 at 16:31
  • @AndrewSteitz look at it this way: the scheme in question makes it harder to uncover a violation, but that doesn't change the fact that there's a violation.
    – phoog
    Commented Sep 9, 2017 at 16:51
  • Exactly! That's why I said I hope it is correct. :-)
    – Andrew Steitz
    Commented Sep 11, 2017 at 16:33
  • 1
    It can not be proven for both, since either XOR1 or XOR2 is actually not derived from the source, one is truly random. That is the idea. I agree that there is a violation, but how can it be clear whether XOR1 was derived from XOR2 and the source, or that XOR2 was derived from XOR1 and the source? Actually, at most one of the hosts of XOR1 or XOR2 may actually be providing a random sequence of data as a service, without knowing that his service was used for copyright infringement. My point is, that by symmetry, it is hard to prove which one started infringing.
    – Herbert
    Commented Sep 14, 2017 at 12:05
  • 1
    My point is that if you can deduce that one of two is guilty, but not which one, can you prosecute either? Can you force the removal of XOR1 or XOR2 or both? Or can the copyright holder claim damages at a person related to hosting either XOR1 or XOR2?
    – Herbert
    Commented Sep 14, 2017 at 15:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .