Technology to circumvent online copyright enforcement
January, 2010
This post could also be called “Why it might become civil disobedience to serve up random data.” You’ll see why in a bit!
Also, this post requires a basic understanding of XOR and how a one-time pad works.
If I serve up Lady GaGa’s latest MP3, I’ll get a takedown notice for copyright violation.
What if a friend and I served up bits that, when XOR’ed together, became the latest Lady GaGa MP3?
A judge would likely find both of us guilty of copyright violation, but it seems like a small stretch of the law. I wonder if there’s something there…
What if I served up bits that, when XOR’ed with some publicly well known bits (e.g. the Firefox installer), became a Lady GaGa MP3?
I would probably be liable; it’d be strange for Firefox to be found guilty in any way. You can see the law now: “you are culpable if you distribute something that is intended to be a reflection of a copyrighted work”. Base guilt on intent and let the court discern intent.
This approach seems reasonable until you imagine it at scale. Any given copyrighted work could be expressed across 10 random-looking files. Are the copyright owners prepared to sue all of them? That’s denial of service for you. But it gets better because some of the hosts might really be innocent; some of the data blocks might be legitimate, for example ciphertext that just looks like random data.
In other words, this system would either (a) circumvent copyright enforcement, or (b) allow me to get arbitrary hosts of random-looking data sued.
Hm. College kids could upload random data to their home directories, just to help out the effort.
I don't think this will ever happen, though it is a fun thought exercise.
Update: There has been a lot of great discussion on this over at Hacker News, in addition to the great comments we've been getting below!
Update 2: This idea isn't intended to indicate that file sharing can suddenly become legal if done in a certain way. It's intended to make enforcement intractable.