This paper discusses the possible consequences of bankruptcy on both exclusive and nonexclusive software licenses; source code escrows and security interests as devices by which a licensee can protect itself in the event of the software licensor’s bankruptcy; and the possible need to register certain transactions involving software with the Copyright Office in order to preserve one’s priority with regard to certain rights against the possible claims of others, particularly a trustee in bankruptcy.

The paper is relatively long and legally technical. Non-attorneys will find it a tough read. Some basic principles of bankruptcy and copyright law are reviewed in the paper, but probably not enough to avoid leaving the layperson bewildered. Nevertheless, the points that are discussed are potentially important to every software licensee. I say potentially, because, despite the authors’ cogent arguments, there are no court cases that yet test their theories.

The authors see the following threats to a software licensee in the event of the licensor’s bankruptcy. If the license is exclusive and recorded in the Copyright Office, the licensee’s right to use may be protected, but the licensor can reject any outstanding obligations under the license, including debugging and maintenance. If the exclusive license is unrecorded, then the trustee in bankruptcy can avoid it altogether. Likewise, the trustee in bankruptcy can reject nonexclusive licenses. Rejection may even terminate the right to use the software.

The authors discuss software escrows, including secured escrows, security interests, and methods of perfecting such interests in licensed software, and how these legal devices may help avoid losing the software if the licensor goes bankrupt, but they also point out dangerous pitfalls in each of these mechanisms. Again, there are no cases as yet to test the effectiveness of such mechanisms.

Several possible inconsistencies in the software copyright deposit and registration requirements are pointed out, particularly those involving unpublished trade secrets. Thus far, however, the requirements appear to have posed no real threat to the validity of copyrights, but the authors raise the possibility of a licensee who has made a good faith effort at registration of a software license by complying with the Copyright Office’s regulations losing the license because a court finds that the regulations do not accord with the Copyright Act. The paper closes with proposed legislative and regulatory solutions to the copyright problems they identify. These include elimination of the registration requirement, the elimination of modification of the deposit requirement, or the identification and indexing of software transactions by transferer and transferee rather than by the title or registration number.

I question only one statement in this complex work. The authors state on page 359 that Section 117 of the Copyright Law allows modification or copying of software for use on a “single machine.” Section 117 speaks of use on “a machine.” Licensors may demand that software be used with a single CPU, but the Copyright Law has no such requirement.