Imagine producing a classic Western without cowboys, saloons, or standoffs. This seems almost inconceivable because these elements are deeply integral to the genre – so much so, in fact, that they are essentially necessary for the creation of such works. Copyright law recognizes and accounts for this, by denying copyright protection to such elements under the “scènes à faire” doctrine. “Scènes à faire” literally means “scenes that must be done.” This doctrine traditionally has been applied in the context of literature and film, to keep classic tropes free for use by artists looking to create works in such genres. The Federal Circuit will soon decide, in Cisco Systems v. Arista, whether the scènes à faire doctrine can also be applied in the context of computer programming, to deny copyright protection for software commands that have become commonplace within the field.
Cisco Systems is a technology conglomerate that is suing Arista Networks, a competing software company, for copyright infringement. Both parties acknowledge that Arista used Cisco’s command-line user interface (“CLI”) program verbatim, including 500 multi-line commands, to configure Arista’s Ethernet switching technologies. At issue, however, is whether the copied material deserves copyright protection. CLIs are computer programs that facilitate communications between the user and computer system by accepting certain pre-programmed text-based commands. The copied material here includes the selection and compilation of text-based commands to be accepted by the CLI, as well as the internal organization of these commands within the program.
District Court Proceedings and Arguments on Appeal
Arista prevailed at the district court level. Applying the scènes à faire doctrine, the Northern District of California found that Arista’s copying was permissible because Arista was “constrained by functionality and preexisting network industry protocols.” That is to say, the Court found that a high degree of consumer familiarity and ease with Cisco’s CLI (as compared with other CLIs) rendered it a necessary element for the creation of Arista’s software. The Court also found that Cisco’s CLI (and the associated commands) had become standard in the industry; the commands had become nothing more than “stock” elements, no more deserving of copyright protection than any stock artistic element traditionally denied protection under scènes à faire. Arista echoes these arguments on appeal.
In its appellate opening brief, Cisco cites evidence that different words and configurations could be employed in CLI commands to perform the same functions, and that Cisco programmers used their own creative judgment in selecting these specific terms and organizational structures. In doing so, Cisco argues that affording copyright protection here would not foreclose all meaningful options for competitors looking to employ CLIs – a main policy concern underlying the scènes à faire doctrine. Through presenting evidence of creative judgment, Cisco further argues that the Cisco CLI was not based on an industry standard at the time that Cisco created it; rather, it had become the industry standard by the time that Arista copied it. Cisco emphasizes that the scènes à faire doctrine only denies protection to compilations that were commonplace at the time of creation (not at the time of copying). Additionally, Cisco contends that nebulous ideas of consumer familiarity and usability are not external constraints sufficient to render the Cisco CLI “necessary” for the creation of Arista’s software. According to Cisco, for a factor to be considered an “external constraint” for the purposes of scènes à faire, the factor must necessarily prevent the copier from pursuing any other options – for example, if the computer system upon which Arista’s software was designed to run were exclusively compatible with the Cisco CLI.
Implications of Potential Outcomes
A victory for Arista would set a precedent for applying the scènes à faire doctrine to work that had become commonplace by the time it was copied, but was original at the time of creation. This could lead to the outcome of transforming certain works into victims of their own success (essentially creating a copyright equivalent of trademark law’s “genericide”). Looked at another way, a victory for Arista would allow greater opportunities for innovation in the software sphere; software developers would be able to build off each other’s programs, free from concerns about copyright infringement.
A victory for Cisco would set a precedent for affording copyright protections to work that is arguably functional in nature. This creates the possibility that inventors of useful computer programs could benefit from intellectual property protection while forgoing the disclosure requirements imposed by patent law. A win for Cisco would also create uncertainty for many existing technologies where certain multi-word commands have become common within the field, and may result in increased cross-licensing and litigation between software companies. We will keep an eye out for the Ninth Circuit’s decision and report back.