The AP v. Everybody: Not So Clearly Fair Use

Today, Michael Arrington made a lot of hay about the AP sending mutliple DMCA takedown notices to the Drudge Retort, a collaborative politics blog. Retort founder Rogers Cadenhead summarizes the matter. To summarize: Users of the Retort copied small snippets (~50 words) of AP content and linked back to the content. The snippets were there to give an introduction to the full article, and also to spark a conversation hosted at the Retort about the facts being discussed by the article. The AP responded by sending takedown notices to the Retort for seven of these alleged infractions.

In response, Arrington, in his well-commented blog post, calls for a boycott on quoting and/or linking to AP stories. Why? Because in part, he claims, the AP “do[es] not want people quoting their stories, despite the fact that such activity very clearly falls within the fair use exception to copyright law.”

But as a professor of mine used to joke, it is almost axiomatic that, when a lawyer or judge makes a naked assertion that a law or legal theory “obviously”, “clearly”, etc., the matter is anything but clear or obvious. That is not to say that the use of the content on the Retort was not fair, but simply that the question is closer — a lot closer — than Arrington and the parade of commentors suggest. In other words, the AP may be wrong on the law, but they may also be right.

Before I continue, let me make clear that I think the AP is acting uncommonly silly by fighting against this tide — they can’t win, and even if they do, they may be doing themselves more harm than good. But that does not mean they are, legally, out of bounds. With that aside, let us look at the four fair use factors.

17 USC 107 sets four fair use factors:

(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;
(2) the nature of the copyrighted work;
(3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
(4) the effect of the use upon the potential market for or value of the copyrighted work.

The first factor is arguably the most important insofar as fair use jurisprudence is concerned. But over the course of decades of judicial gloss, really has little to do with the “commerical” or “nonprofit” dichotomy. Now, the question focuses more on to whether the use is “transformative”. The Stanford Copyright & Fair Use page asks two questions to help determine if a use is transformative

(a) Has the material you have taken from the original work been transformed by adding new expression or meaning?
(b) Was value added to the original by creating new information, new aesthetics, new insights and understandings?

I can’t imagine either is the case. Contrast the use on the Drudge Retort itself with, ironically, the use on Cadenhead’s wrapup blog post. The latter one discusses the allegedly infringing use in the context of discussing how uncommonly silly the AP is being. The words are given new meaning and new insights are being drawn, as the value of the words is to discuss the validity of the AP’s claim. On the other hand, the original copying of the words lacks transformation — the author was, mostly if not entirely, simply relaying the news as discussed in the AP’s article. That was what the AP was trying to do, to.

In my book, this factor goes to the AP.

Number two — “nature” — deals mostly with whether the copied work is factual in nature. This goes to the Drudge Retort submitter, as the AP is reporting (we hope) the facts of the matters discussed in their articles. Note that this factor not dispositive, as “facts” and “content which is factual in nature” are not synoynms. Facts in and of themselves earn the author no copyright protection, but the expression of facts does deserve protection de jure

I’m going to skip #3 for a bit and turn to #4. So bear with me for a moment.

The fourth factor asks whether the use “deprives the copyright owner of income or undermines a new or potential market for the copyrighted work” per Stanford. This one should be easy — AP, hands down. First, local newspapers often use snippets of AP articles in lieu of the full article — sometimes, as few as 50 words. (If you are in New York, pick up an amNY or Metro near a subway stop. You’ll see 10+ per issue.) Second, the AP could very easily — and probably should — create it’s own Digg/Mixx/Retort/etc. for AP stories.

Unfortunately, some bloggers flip the test on its head, stating that “little quotations posted by bloggers are not stifling demand for the AP’s product. Bloggers are creating demand, not decreasing demand by creating replacement supply.” That is a great point and I would guess that it is also the reason why Arrington et al are rightfully up in arms. But it does not help the Retort’s case as much as my argument above hurts the Retort. The AP would likely argue that the blogger quoted above is simply wrong, and that the Retort’s use does, indeed, stifle demand for the AP’s product. After all, why should the small-town newspaper or other AP subscriber pay for the content that the Retort uses for free? The AP has every right in the world to limit the supply — that’s exactly the point of the monopoly power of copyright. They do so, in theory, to increase revenues. While they may be doing a bad job there, I can’t imagine that the Retort would be able to use that fact as a way to triumph on factor four. So, this one goes to the AP.

So let’s hit the third factor.

The analysis under the third factor is both quantitivative and qualitative. Harper & Row Publishers, Inc. v. Nation Enterprises, 471 U.S. 539, 105 S.Ct. 2218 (1985). Harper is not directly on point, but the tenet of law holds — if you use the “heart” of the content, this factor is not going to favor you.

The quantitative analysis helps the Retort — at most, the clips were 79 words long. No argument there. But whether the Retort users took the “heart” of the content is up in the air. Here’s the test I propose: Let’s say that there are two Retort posts dealing with the same story. One copies the AP headline verbatim and adds no other description. The other uses the first 50 words of the article as well. Which does more page views? Which does more click-thrus? Which has more comments? Using these three metrics, we should be able to tell if the user hit the heart. If the latter does more page views and comments but proportionally dramatically fewer clicks, that is good indicator that the “heart” is there on display. After all, why bother reading the whole thing if you already read the good stuff?

My formula may be off — but that’s OK. The point is that the third factor is more in play than we would facially believe. So unless one is willing to adopt a bright-line rule that a quantitatively small uses of content which is factual in nature is an ipso facto fair use, the AP has a claim. It’s not necessarily a winner, and it’s foolish for them to pursue it, sure. But it’s there nonetheless.

Dan Lewis is a co-founder of ArmchairGM.com and now works at Wikia. He also blogs at DLewis.net. You can also subscribe to his RSS feed.