RLI 279 20 June 2012 Research Library Issues: A QuarterlLy Report from ARL, CNI, and SPARC Copyright Risk Management garner few complaints, but in the rare circumstances where a family member objects that an ancestor’s letters, for example, are being displayed to the world, they will often be mollified if the material is removed from public view and the objector is invited to discuss the matter. Sometimes these discussions may result in eventually reposting the work(s) in question, in paying a small licensing fee, or in deciding to leave the material out of the collection. But a responsive take-down policy will inevitably have the effect of preventing most complaints from ever becoming lawsuits. It is important to note that a take-down policy in the context of library-created digital collections does not have a legal status it does not create the “safe harbor” that the take-down process outlined in the Copyright Act offers to Internet service providers.8 There is therefore no guarantee that a rights holder could not or would not sue for an alleged infringement even after the offending materials were removed. But from a risk-management perspective, this is an effective way to defuse conflicts if they arise and will further reduce the anxiety around a digitization project. The final strategy in this arsenal of risk-management techniques is, of course, recognizing that many collections will be supported by a strong fair-use argument. While it is not necessary to rehearse all of the details about what such an argument would look like, two points are important. First, most of the mixed digital collections of 20th-century material to which these strategies would apply will clearly be transformative such collections will repurpose the individual materials around a research theme, in most cases, that will be far different than the original purpose of the works and will not in any way compete with that original purpose. Second, there is unlikely to be any market for the original in many of these cases, even if licensing markets are taken into account. So the two arguably most important fair-use factors, the first and fourth, will often favor the creation of these digital collections. When we discuss fair use it is import to consider its application and impact on digital collections decision making. In a mixed and heterogeneous collection, any fair-use argument fabricated in advance of a specific complaint will not apply equally well to all materials. But as we have said, this approach does not require absolute certainty or universal application. When the goal is to evaluate the level of risk in order to undertake sensible digitization projects, it is enough to recognize that fair use would be a plausible defense, and that a good-faith fair-use defense reduces the availability of damages when the user is a nonprofit educational institution.9 This is especially the case where fair use is understood to be a part of a wider strategy a “last line of defense” that would further deter potential plaintiffs if all of the other strategies proved, in some rare instance, to have failed to prevent a complaint. For that limited pool of material that is neither in the public domain nor subject to permission, and for that rare plaintiff who is not satisfied by a take-down process, fair use still provides a boundary to the copyright, and raising it would increase the probability that the plaintiff would decide that a lawsuit was likely to prove too expensive and too unprofitable to undertake. Applications This four-prong strategy has been successfully applied to two projects in which I have been involved. In the case of a collection of historic TV commercials, recognition of the public domain and efforts to obtain permission from major rights holders were instrumental in a decision to proceed with the digitization.10 The fair-use argument provided a kind of “backstop,” especially in a couple of instances where a putative rights holder told the library that they were not comfortable giving the asked-for permission. In those