The Eleventh Circuit case that struck down mandatory metal detectors for protest attendees (cited by Orin Kerr below) is noteworthy for one reason besides its important and likely controversial holding: It cites Wikipedia, a free online collaborative encyclopedia, for information on the Department of Homeland Security Advisory System. It is one of only two cases I've found that cite Wikipedia, the other being Bryant v. Oakpointe Villa Nursing Ctr., 471 Mich. 411 (2004), a Michigan Supreme Court case, which cites Wikipedia for information on positional asphyxia.
Now I much admire the Wikipedia project, and my hat would be off to Larry Sanger and Jimmy Wales, its cofounders, if I wore a hat. The concept of an encyclopedia that is cowritten by lots of people, each of whom has the power to edit any of its pages — with the main screening mechanism being the possibility of correction by others — sounds odd. But it seems to work pretty well; and of course the real question isn't whether the work is perfectly reliable, but (1) how reliable it is compared to the alternatives, (2) whether that's good enough for the particular use you're making of it (e.g., casual attempts to satisfy curiosity rather than decisions where someone's life or even a lot of money is on the line), and (3) whether the work's advantages in thoroughness, currency, convenience, and low cost exceed the possible reliability disadvantages. (Here, by the way, is the Wikipedia response to the arguments that free editing may make the encyclopedia too unreliable.)
Still, I wonder whether it's good for court opinions, which not only resolve disputes between parties but also effectively create law that governs future disputes, to rely on something that at least has the potential to be so easily compromised, whether as part of a deliberate strategy or not. Of course, court opinions can likewise screw up by citing to the many erroneous portions of Michael Bellesiles' work (which some have done); seemingly reputable work by a noted professor or by an established reference-work-producing organization may be mistaken, too. And I suspect the main source of error in court opinions isn't relying on simply mistaken information but rather relying on one source that says one thing when a dozen other more reliable sources that the court hasn't found say the opposite, and more persuasively. Maybe on balance Wikipedia is good enough, especially when the information that the court is drawing from it is likely to be pretty uncontroversial. Nonetheless, it strikes me as something that judges and law clerks should be cautious about using.
Incidentally, a quick search through WESTLAW's BRIEFS database, which contains a subset (but likely a very small subset) of all recent briefs filed in appellate court cases, reveals two briefs where the lawyers likewise cited Wikipedia. On the one hand, that's especially dangerous for a lawyer: If a judge or the judge's clerk knows about Wikipedia's collaborative model, and therefore doubts Wikipedia's reliability, then the citation may (fairly or not) undermine the brief's credibility. And that's especially so if the judge or the clerk looks into it more closely and finds that the Wikipedia article is indeed mistaken.
On the other hand, lawyers, even more than appellate judges, have to work quickly, and often have to cut corners on research. So maybe for some fairly tangential point, a cite to Wikipedia might be seen as the most efficient solution (though subject to the credibility loss concerns that I mention above, which may be present even if the Wikipedia entry proves to be entirely accurate).