With a new round of litigation threats from the Recording Industry of America, there has been a lot of attention, and confusion, about what obligations internet service providers have to respond when notified of allegedly infringing activity. The phrase “take-down notice” is beginning to enter our collective vocabulary, but it is important to understand to which situations it does, or does not, apply.

The take down notice is a product of section 512(c) of the Copyright Act, added by the Digital Millenium Copyright Act of 1998, and it is applicable only when the allegedly infringing material resides on the servers owned by the internet service provider. There is substantial evidence that these notices are used by content owners to remove critical material from the web even when the claim of copyright infringement is very weak.

A recent, almost comic, contratemps between law professor Wendy Seltzer and the NFL nicely illustrates how take-down notices work. Prof. Seltzer copied and uploaded to YouTube a short section of the Super Bowl broadcast — the part where they read that overly broad copyright warning that says that even “descriptions and accounts” of the game must be authorized by the NFL — in order to illustrate the absurd claims that some copyright owners use to try to frighten people from making legitimate uses of content. The NFL promptly sent a take-down notice to YouTube, which removed the video and notified Professor Seltzer. She responded with the statutory counter-notification that asserted her posting of the video clip was fair use, and YouTube put the video back. Then the whole cycle was repeated again. The clip, restored for the second time on April 4, is still on YouTube as of today, and Prof. Seltzer has certainly made her point about how copyright assertions, and especially the DMCA take-down procedure, can have a chilling effect on legitimate expression. (See Wendy Seltzer’s blog as well as the “Chilling Effect Clearinghouse“).

But when infringing material does not reside on the service providers’ own equipment (as clips in YouTube reside on Google-owned servers), then neither the take-down notice and counter-notice nor the expedited subpoena process apply. Thus, in cases of peer-to-peer file sharing, where the infringing material is only transmitted over the service provider’s network but is not stored or maintained on that network, the 512(c) procedure can not be used.

So what is the RIAA sending to universities regarding file sharing? The recent batch of letters sent to university DMCA agents, who are designated to receive take-down notices, have taken two forms. One is a “settlement offer” that is identified only by an IP address and that asks the institution to pass on to the student associated with that address. When and if the letter is passed on to the correct student, that person is told to go to a specified website and settle the claim for infringement or risk being sued. Since this procedure saves the RIAA the time and trouble of getting a subpoena to learn the identity of each student alleged to be sharing files, it is much more efficient for the RIAA; it puts the institution in the middle instead.

Some schools have responding by saying that they do not retain server logs long enough to match the dynamic IP addresses referenced in these settlement letters to the offending student, so the RIAA has added another letter, demanding that the institutions retain records in anticipation of possible litigation. Whether or not this demand is legally enforcable is a debated issue, but many universities are complying with both letters out of concern not to leave students unaware of their risk of litigation. So now the “chilling effect” that has long been associated with the DMCA is being exploited to save time and increase settlement revenues for the recording industry.

There has been lots of talk in copyright arenas recently about settlements. Some settlements, like the “pre-litigation” settlements the RIAA is offering to college students, where the industry collects roughly the same amount they had been getting from their lawsuits without having to bear the expense of actually filing, are clearly bad deals for the public.  But the downside of other settlements is not so obvious.

Recently English professor Carol Schloss settled her lawsuit against the James Joyce estate. She had sought a declaratory judgment that quoting from unpublished letters between Joyce and his daughter Lucia in a scholarly book was fair use, and the settlement was clearly a victory for Schloss. The Joyce estate agreed not to sue her for using the excerpts, which had been removed from her book about Lucia under threat of an infringement claim, on her web site and in future publications. But it is less clear that this is a victory for scholarship in general. If the case had gone to trial, a decision would have had precedential value, and we would have had another bit of clarity in the murky world of academic fair use. Fair use is so ambiguous that scholars often refrain from certain uses simply to avoid the uncertainty. Most of us assume that quoting from primary sources in scholarly work, just as Schloss wanted to do, is fair use, but it would have been nice for a court to confirm our impression.

Instead, the settlement agreement only resolves the issue between the parties to the case and has no impact on fair use for the rest of academia. In fact, the AP paraphrased the attorney for the Joyce estate this way: “Nelson also noted that the estate granted only Schloss permission to quote the materials under limited circumstances, meaning neither she nor other scholars would be permitted to use them outside the scope outlined in the settlement agreement.” Not really a Phyrric victory, but not a helpful one either.

Suggestions that the various publishers who are suing Google over the Book Project might settle for some lump sum payment from Google’s deep pockets offer a similar prospect. If Google were to lose its fair use claim in court, of course, the effect on fair use for the rest of us could be profoundly negative. But if they fought on and won, the prospects for library digitization and public access would be much brighter.  A settlement, however, will set only one precedent — that lots of money can solve all disputes. It will not clear up the fair use picture and it will not benefit libraries that can not afford to purchase the right to digitize their collections for the benefit of the public.

So with all due respect to my grandmother’s admonition to get along with everyone, I have to recognize that sometimes there is a downside to playing nice and coming to a settlement agreement.

When Viacom filed a copyright infringement lawsuit against YouTube, which is owned by Google, we all knew that the fur would fly. No one is neutral about Google these days, and YouTube has also created fierce partisans in the copyright wars.

It is quite possible that the lawsuit is simply a bargaining tool in the fight over the licensing payments that Viacom wants from Google, but, if the case goes to trial, the real issue will be how far copyright law can be used to prevent technological innovations that depend, to some degree or another, on using copyrighted material without permission. This debate is at least as old as the photocopier and the VCR, but the terms of the rough deal stuck by the Supreme Court in the 1984 Betamax case are still being contested.

Several newspaper and online columns over the weekend cast interesting light on this debate .

In the New York Times on Sunday (March 18), Larry Lessig of Stanford Law School makes the important point that Viacom is trying to use the courts to get around the compromise on innovation versus infringement that was established by Congress when it passed the Digital Millennium Copyright Act. The DMCA created a “safe harbor” for ISPs that merely hosted infringing material on their servers without specific knowledge of the infringement, so long as they act according to prescribed procedures once they have such knowledge. Now Viacom, emboldened by the Supreme Court decision in the Grokster case, is trying to revise that compromise through the back door, using the Courts new “inducement” theory of secondary copyright liability to undermine the policy Congress enacted.

Many predicted that the Grokster decision would have a chilliing effect on technological innovation; Lessig’s column suggests that the Viacom suit is a huge block of ice to contribute to that chill.

Along the same lines, Fred von Lohmann of the Electronic Frontier Foundation writes about all of the new technologies that might not have seen the light of day if copyright holders could sue based on the potential that users of the new innovation could infringe those copyrights. Even if one excludes the hyperbolic inclusion of reading glasses on his list, von Lohmann makes a powerful point — photocopiers, VCRs, iPods and even public libraries might be subject to restrictions and per-use fees in the world Viacom envisions.

James Boyle, in a column in the Financial Times, tries to put this debate in perspective by noting that, although it is sometimes difficult to feel sympathy for the ravenous giant that Google has become — it is no longer, as Boyle says, “a helpless start-up” — the interests that Google has pursued in its copyright conflicts often line up with the public interest. His reminder that the issues are much more complex than simply yelling “stop thief” at YouTube and Google (a tactic which Google’s rival Microsoft tried in a recent speech and op-ed), and that we as consumers of intellectual property also have an important horse in this race, are welcome indeed.