Ars Technica has an article about whether iTunes’ new iCloud features will be used to chase down music pirates. Some lawyers and academics writing about this might hedge their comments with conditions and exceptions. I won’t.
No, iTunes’ new iCloud match features cannot and will not be used to chase down music pirates. (And Ars Technica’s conclusion is correct.)
First, let’s reject the pirate label
My post on why you can’t steal an idea touches on why ‘music pirate’ is the wrong term. You can’t steal music—you can only violate someone’s copyright. These are not the same thing. Nina Paley sums this up nicely in a song she wrote about why copying is not theft; something I touched on in my post about the difference between rivalrous and non-rivarlous ‘things.’
Copying music you don’t have a right to copy isn’t theft—it’s a copyright violation. You’re not a ‘music pirate,’ despite how the Recording Industry might want to paint you (or you might want to paint yourself, because Johnny Depp made pirates cool again).
Second, what’s the argument iTunes will lead to a mass copyright violation hunt?
The Ars Technica article asks two questions: (1) “Could Apple reliably identify pirated music?” and (2) “What would [Apple] do with that information?” It cites a recording industry insider answering the first question as yes, Apple will be able to identify pirated music.
Another article, cited by both Slashdot and Techdirt, goes further and argues that iTunes’ music match service might be a ‘honey pot’ for identifying copyright violators. In this Between the Numbers article, Daniel Nolte lays out the argument that Apple will be able to identify individual files that have been ‘stolen’—or in legal terms, copied without permission (and therefore a copyright violation).
The gist of these arguments is that meta data contained in an MP3 file on a computer might allow Apple or a music company to identify that file as one downloaded or copied illegally. MP3 files can contain all sorts of meta data—from the username or email of the person using an MP3 download service like Amazon’s, or a hash tag identifying that download specifically to a purchase by someone, or any other identifying features. Both the Ars Technica article and the Between the Numbers article detail how this might work.
Third, why these arguments are legally rubbish
Each of these arguments focus on the uniqueness of Apple’s service and the potential information on users it might provide, while ignoring the realities of what it takes to make a prima facie case for copyright infringement. Even presuming each of the technical arguments made by folks in both articles—which can only be a presumption, as the Ars Technica article rightly points out we do not know how Apple’s service actually works—a specific song cannot and will not be identifiable as infringing another’s copyright without additional information.
So what is this ‘additional information’ you mention?
A prima facie case for copyright infringement is the bare minimum that must be alleged to prove copyright infringement if all the factual statements are taken to be true. Section 501(a) of the Copyright Act of 1978, as amended, states that a copyright infringer is “[a]nyone who violates any of the exclusive rights of the copyright owner as provided by sections 106 through 122.”1 Section 106 of the Copyright Act of 1978, as amended, lists two exclusive rights we’re interested in for the purposes of music sharing: (1) reproduction and (2) distribution.2
So a prima facie case for copyright infringement is one where the defendant has copied or distributed a copyrighted work (song) without the permission of the copyright’s holder (the music company, et cetera). At first blush, the information Apple is likely to collect (according to the articles above) will include the information necessary to make this claim. Except it doesn’t.
The idea of the tech arguments in the above articles is that the unique identifier Amazon has in the meta data of the MP3 can identify my MP3. And it could. And it might show where other accounts have that MP3 on Apple’s service. Ah ha, some might say, evidence of infringement! And they’d be right.
Except it’s not enough evidence to identify who is the infringer, and which file or files are the infringing files. Perhaps I have two iTunes accounts (which I actually do, for various reasons, and so do others). Perhaps it was uploaded twice or more as a result of that. Perhaps a family member copied it without my permission and against my knowledge. Perhaps I violated copyright law and copied it and distributed it to anyone I know.
The point is we don’t know which scenarios above apply based on the limited evidence Apple’s data presents. All we would know is that there are duplicates of a file where there should only be a few (or one). In this situation all the infringing copies are scooped together in the data with all the non-infringing copies.
Compare this with the data you’ve read about in the file sharing cases that have been ongoing. Each of those cases identified an individual who was sharing the data. The evidence culled from file sharing services had that person sharing music that was copyrighted. This creates a prima facie case for distribution of a copyright work without permission.
In the Apple iTunes match situation, there is no distribution without permission. Apple’s received the permission from recording companies for you to distribute and copy MP3s to its service. (In fact, in a situation where music is identified and its in the iTunes catalog, you’ll be using Apple’s copy.)
It’s not like the file sharing cases that have caught headlines—no one is caught red handed (and no one can be caught red handed). At most there is evidence that a file might have been shared illegally, but no evidence exist to point to who violated the law.
What about my ripped CDs I upload to Apple?
Ripped CDs are even less problematic. Unlike MP3s downloaded from Amazon or elsewhere, ripped CDs only contain metadata included during the copying process. There are no unique identifiers as their might be in an MP3 you download from Amazon (or Apple). So even the idea that a unique identifier can be used to identify illegally copied music is moot.
Ripped CDs might provide evidence of copyright infringement, but only on a statistical level. For example, if only 100 CDs of an album have been bought, but Apple’s users collectively have 1,000 of them in the iCloud service, it’s clear unauthorized copying has occurred somewhere. But, once again, and as in the above situations, no evidence exists to show who was the illegal copier.
Lawyers and legal scholars worrying about this should calm down and reexamine copyright law
I’ve come across a number of discussions of this conspiracy theory—that Apple is (or might because of legal pressure) sell out its iCloud and iTunes Match users to copyright holders—and I am unimpressed with the lack of analysis. Ars Technica’s article reaches the right conclusion, even if it hedges more than this post.
The legal scholars Ars Technica’s writer Chris Foresman interviews for the article are on point. Julie Samuels, a staff attorney for the EFF, points out in Foresman’s article that the grounds for getting a subpoena for Apple’s data is will be difficult to show. Evan Brown, a Chicago-based attorney, points out in Foresman’s article that the practicality of getting Apple’s data is limited because what is in that data isn’t publicly available.
The reason, dear readers, is that Apple’s data in-and-of-itself is insufficient to sustain a claim for copyright infringement. More information than Apple can and will collect is necessary, and that’s why Apple’s new iTunes Match service cannot and will not be used to chase down ‘music pirates’ (or more accurately, copyright infringers).