I blogged about this two years ago, and concluded that the site would and should win under the fair use doctrine. Yesterday, the Fourth Circuit accepted the site's fair use claims. Here's my summary of the matter from last year, though you might instead just read the Fourth Circuit opinion:

Turnitin.com is a commercial service that aims to help educators catch plagiarism in student papers. Schools require that student papers be submitted to the site, which (1) checks each student paper against its database, and (2) adds each student paper to its database so that future papers can be checked against it. I take it that the database already contains papers from commercial term paper mills, encyclopedia entries, and the like; but adding student papers helps spot students who are copying from classmates, or from friends at other schools, as well as students who are copying from publicly available sources.

But, the high school student plaintiffs say, step 2 violates our copyright: You folks are making money by copying our papers onto your servers. The consent you get from us is inadequate because we are coerced to give it (especially plausible, I take it, when the students are students at public high schools, and when they are within the compulsory school attendance age range). And your use is not fair use, chiefly because it's commercial....

Turnitin.com's strongest fair use argument (which would be needed if the court concludes that the student did not voluntarily consent to the use of their works) is that though their use is commercial, it is

1. transformative — it copies the papers not to use them as papers (as opposed to, say, a Napster user, who copies a song to play it as a song), but rather to use them to check other papers for plagiarism --

2. is in aid of others' nonprofit educational mission, and

3. does not interfere with the students' market for their own works, since the students' works are worthless, and in any event if they are worth something (say, because the students can sell them as newspaper op-eds or articles in literary magazines), Turnitin's archiving of the papers wouldn't interfere with that value.

The students' strongest response would essentially be that if Turnitin is making money from the students' works, the students are entitled to a share of that, and Turnitin's using the works for free interferes with the students' ability to license their papers to Turnitin itself. The most familiar analogy here would be if Steven Spielberg decides to make a movie out of your novel. It may well be that the movie won't interfere with the value of your novel -- it may even increase your sales--— but his making the movie without paying you interferes with your right to license movie rights to the novel. (Some condemn this as circular reasoning, but I don't think that's quite right, and in any event it is precisely the reasoning that lets authors profit from selling movie rights to their books, and that bars moviemakers from just using the books for free and claiming fair use.) If anything, the students would say, our case is even stronger because our works are unpublished, and the unpublished status of a work is generally seen as cutting in favor of the work's owner and against the fair use claims of the user.

[But] I'm pretty sure Turnitin would and should win, because (1) the value of the licensing rights in their papers would in any case be next to nil, (2) Turnitin's use is transformative -- in the sense that it uses the original to make something that's very different (much as a parody or a photo search engine that presents thumbnails of others' photos is transformative, though not quite in the same way) -- and therefore is not within the authors' legitimate licensing rights (cf. the Supreme Court's holding that "there is no protectible derivative market for criticism"), and (3) the unpublished status of plaintiff's work should only matter when the defendant is trying to publish the work (or a version of the work), which it isn't doing.