Wednesday, September 22, 2010

Vernor v. Autodesk Three-Pronged Ownership Test Is Fatally Flawed

On appeal in Vernor v. Autodesk, the court defined a new three-pronged test for determining whether a user of software is an "owner" or merely a "licensee" within the meaning of sections 109 and 117 of the Copyright Act. It is my opinion that this new three-pronged test is fatally flawed.

First, let's introduce the three-pronged test:

a software user is a licensee rather than an owner of a copy where the copyright owner (1) specifies that the user is granted a license; (2) significantly restricts the user’s ability to transfer the software; and (3) imposes notable use restrictions.

Now let's apply this new test to some hypothetical cases. In doing so we'll demonstrate how the test fails to correctly determine whether or not the user is the owner of the copy. I've chosen these particular hypothetical cases because it is otherwise quite clear from the facts of the cases whether the user is or is not the owner of the copy. The first hypothetical case shows that application of the three-pronged test can lead to the incorrect conclusion that a user is an owner of a particular copy:

Alice asks Bob if she can use a copy of a program Bob wrote. Bob copies the program to his own USB flash drive and loans it to Alice. A week later, Bob asks to have the flash drive back, but Alice says she sold the USB drive to Charlie. Bob sues Alice for infringing his exclusive right to distribute copies of his program. Alice asserts she had a right to sell the copy under the first-sale doctrine.

The court must now decide whether or not Alice was an "owner of a copy" in order to determine if she can claim protection under the first-sale doctrine. Applying the three-pronged test:

  1. Bob did not grant a license to Alice;

  2. Bob did not specifically restrict Alice's ability to transfer the program; and

  3. Bob did not impose any use restrictions.

It is clear from the facts that Alice is not an "owner of a copy". The USB drive, which is Bob's property, is the copy. Bob merely loaned the USB drive to Alice; he never transferred ownership of it to her. However, according to the three-pronged test, Alice clearly was an "owner of a copy" and therefore had a right, under the first-sale doctrine, to sell the copy to Charlie. The test fails to recognize that Bob was always the owner of the USB drive. And since the USB drive is the copy, Bob was always the owner of the copy. The fact that Bob didn't extend any license to any of Bob's copyrights simply has no bearing on who owns the USB drive and, hence, who owns the copy. If we stick to the three-pronged test, we must somehow explain how ownership of the USB drive was implicitly transferred to Alice (even though Bob's and Alice's original agreement was that the USB drive was merely loaned to Alice).

In the next hypothetical case, we demonstrate that the opposite outcome is also possible. That is, application of the three-pronged test may lead to the incorrect conclusion that a user is not the owner of a particular copy. The facts in this case are similar to the first, but this time the program is copied to Alice's USB drive and Bob additionally grants Alice a limited license:

Alice asks Bob if she can use a copy of a program Bob wrote. Bob copies the program to Alice's USB flash drive. Bob tells Alice he grants her permission to make more copies of the program, provided she doesn't give any copies to anyone else and provided she promises to only use the program for working on a specific school project. Alice never makes any other copies of the program. Later, she sells her USB drive to Charlie. Bob finds out about the sale and sues Alice for contributory copyright infringement. Alice again claims she is protected by the first-sale doctrine.

Once again, the court must determine whether or not Alice was an "owner of a copy" in order to determine if she can claim protection under the first-sale doctrine. Applying the three-pronged test:

  1. Bob granted Alice a license;

  2. Bob's license restricts Alice's ability to transfer the software; and

  3. Bob's license imposes significant use restrictions.

It is clear from the facts that Alice is the owner of the copy -- the USB drive belongs to her. However, according to the test, Alice was merely a "licensee" and not an "owner of a copy". The test fails to recognize that Alice was always the owner of the USB drive. And, again, since the USB drive is the copy, Alice was always the owner of the copy. It also fails to recognize that Alice was both an owner and a licensee. If we stick to the three-pronged test, we must somehow explain how ownership of the USB drive automatically transferred to Bob once his program was copied to it.

The reason the test fails is three-fold:

  1. The test relies on a logical fallacy known as "begging the question": the premise of the test, that the terms in the license are valid and binding upon the user (i.e. "the user is merely a licensee"), depends on the truth of the very matter in question (i.e. "is the user merely a licensee?").

  2. There is an implicit and incorrect assumption within the test that a user can be either an "owner" or a "licensee", but not both (if you are a "licensee", then you are not an "owner of a copy").

  3. The test assumes that ownership of a material object can be determined solely by looking to the terms of a license of intangible rights which may be completely unrelated to the transaction in which the user obtained possession, and possibly ownership, of the material object.

Because ownership of a copy of software is inseparably tied to ownership of the material object in which the software is fixed (i.e. the media it is recorded on), the question of ownership of the copy must look at how the user came into possession of the material object and whether or not the transfer of possession also transferred ownership. Any licenses of copyrights are entirely orthogonal to the question of ownership of the material object. In our second hypothetical case, Alice clearly owns her USB drive; the terms of Bob's license agreement allowing her to make additional copies is simply irrelevant -- yet the Vernor court's test would have us rely solely on the content of Bob's irrelevant license agreement.

In summary, application of the Vernor v. Autodesk three-pronged ownership test can be shown to reach the wrong result. The test fails because:

  • it is based on a logical fallacy;

  • it doesn't recognize that licensees can simultaneously be owners of copies; and

  • it confuses the act of licensing intangible rights with the transfer of ownership of material objects.

Tuesday, July 27, 2010

The Third Chamber of Congress

No too long ago, the Software & Information Industry Association (SIIA) wrote an amicus brief in support of Blizzard in the infamous MDY v. Blizzard case. Within that brief, the SIIA wrote:

Today, licenses govern most mass market software transactions, and the software industry has grown to over $500 billion in annual revenues.

Here the SIIA points out that the $500 billion software industry almost exclusively uses so-called "licenses" to distribute their products. The appeals court will hopefully carefully take note of this fact, because it is an indicator of what will happen to our country's copyright laws if the district court's MDY v. Blizzard decision is upheld.

The thing is that section 117 of the Copyright Act only has meaning if users of software own the copies of software that they use. If the copyright owners own all of the copies, then there is nobody who is affected by section 117. Section 117 would basically become a dead law. A ghost-town within the law books, complete with chirping crickets and tumbleweeds. It might as well just be dropped from the books.

In other words, by allowing the SIIA and the companies it represents to unilaterally claim that all software is licensed, not sold, the courts would effectively be promoting the software industry to the position of the Third Chamber of Congress. It would give the software industry the ability to write their own "laws" about what actions (or inactions!) would constitute copyright infringement anywhere software is involved. This is because, without section 117, EULAs would be the only voice that would determine what is and what is not copyright infringement. Imagine that a company would get to decide what rules consumers will need to follow, lest they be hammered by the full force of copyright law complete with its exorbitant statutory damages and even criminal penalties. This is a pretty scary prospect, and is a prospect that the appeals court should not take lightly when considering MDY v. Blizzard.

Thoughts on MDY v. Blizzard

A commenter recently asked what I thought about the MDY v. Blizzard case. For the most part, I think the court makes the same mistakes that we've seen in other cases. The court surmises that simply because the EULA states that the software is licensed, then it must be true. Ergo, they say, the user is a licensee, not an owner of the software and section 117 does not apply.

According to the court, the only thing that makes the transfer of a copy of WoW a license instead of a sale, is the EULA itself:

The EULA thus makes clear that Blizzard is granting to its users a license, not ownership, of the copies of the game client software.
The court ignores the fact that by all other accounts, the transaction would perfectly fit the definition of a sale. And what if the buyer rejects the EULA? Then who owns the copy? The EULA doesn't apply if it's rejected, so any requirement within it to return the software is inoperable. It seems to me that the end user would own the copy of WoW in such a case -- then what?

The bottom line in this case is the same as it is everywhere else. The question of ownership turns entirely on who owns the media on which the software resides. In their amicus brief in support of Autodesk in a separate, but similar case, the SIIA refers to the media as, "a worthless plastic CD." This is very telling of the software industry's indifference to who owns the media. And the reason for their visible indifference is because they know they can't demonstrate ownership of the media. After all, what owner in their right mind would allow the property they own to be distributed all around the country, not knowing who possesses it or where it's located? Yet, this is exactly what software companies do.

But why is ownership of the media so key? It's because the media is the copy. The Copyright Act specifically defines it as such. Whoever owns the media, owns the copy. It's really as simple as that. Bought software online? The copy resides on your hard drive? You own the hard drive, you also own the copy. Same goes for the owner of a CD-ROM or floppy disk.

In the case of WoW, it's pretty clear that the users own the media. Again, Blizzard has no idea who possesses the media or where the media is located. It would be quite a stretch for them to claim they own the media. It's even possible to buy WoW as a direct download: there is no media for Blizzard to own, only the user's hard drive. Any claim that Blizzard owns the media (especially if the media is your own hard drive!) is plain nuts, and patently false.

I also find this sentence from the court's opening statements somewhat amusing (emphasis added):

A user can obtain the game client software by purchasing a copy at a retail store or downloading a copy from the WoW website.

Apparently, even the court is still confused as to whether users are licensing or purchasing the copies. I find it telling that they would make such a mistake. I believe the reason for the mistake is because it feels natural to characterize these transactions as purchases/sales, because they seem like it in virtually every way. Even the court can't shake the feeling.

In my opinion, the court was mistaken when it ruled that MDY committed contributory copyright infringement. However, there's more to the MDY case than copyright infringement. There are also questions related to the WoW terms of use (TOU). The TOU of the online services are governed by a completely different set of rules. It seems likely that MDY did cause end users to violate the TOU, and MDY does admit as much.