In all copyright infringement cases, a plaintiff must prove, among other things, that the defendant copied elements of plaintiff’s work that are protected by copyright. This is key because not all copying is infringement – some of what is copied may be merely ideas, processes, facts, in the public domain, or scenes a faire material. It’s not illegal to copy those things. So a successful plaintiff has to show more than “copying in fact”. It must show “illegal copying”.
Software infringement cases present some nuance for this analysis. A computer program has different levels of abstraction (i.e. from main purpose down to object code), and when the only similarities are at higher levels of abstraction, there is less chance that infringement has occurred. Some courts employ the “abstraction-filtration-comparison test” to evaluate whether a defendant accused of infringing the copyright in software has indeed illegally copied protected elements of the plaintiff’s work:
At the abstraction step, we separate the ideas (and basic utilitarian functions), which are not protectable, from the particular expression of the work. Then, we filter out the nonprotectable components of the product from the original expression. Finally, we compare the remaining protected elements to the allegedly copied work to determine if the two works are substantially similar.
Plaintiff sued one of its founders for copyright infringement after that founder had moved to another company and had developed software allegedly similar to software he had created while at plaintiff-company. The parties agreed to have a special master evaluate the parties’ software to opine on whether defendant had infringed. The special master found there to be infringement, and the district court agreed, ordering that copies of defendant’s software be destroyed.
Defendant sought review in the Tenth Circuit. On appeal, the court vacated and remanded. It held that the special master failed to properly document the steps involved in conducting the abstraction-filtration-comparison test.
The court found there was little evidence the special master performed the abstraction step. Although “[a]pplication of the abstractions test will necessarily vary from case-to-case and program-to-program,” a “computer program can often be parsed into at least six levels of generally declining abstraction: (i) the main purpose, (ii) the program structure or architecture, (iii) modules, (iv) algorithms and data structures, (v) source code, and (vi) object code.” In the court’s mind, “[a]bstraction is important, and it cannot be neglected here.”
The failure to “conscientiously and systematically” perform the abstraction step tainted the remainder of the three-part test. The court criticized the special master’s application of the filtration and comparison steps, observing that the special master apparently proceeded from the false premise that an infringement analysis begins and ends with “copying in fact.” The special master went to great lengths to show that defendant took steps to conceal his copying of source code (e.g., by omitting comments). But having not first properly separated out (by filtering) the unprotected elements after abstraction, the special master’s report was not sturdy enough to support a finding in the district court that infringement had occurred.
Paycom Payroll, LLC v. Richison, — F.3d —, 2014 WL 3377679 (10th Cir. July 11, 2014)