A conundrum employers face when protecting trade secrets is the obligation to identify the very secrets to be protected in litigation.  A critical issue in the context of discovery is the extent to which trade secrets must be disclosed and identified by the plaintiff.  When the secret is something as technical and unique as a computer source code alleged to have been used to augment a competitor’s source code library, the requirement to identify the misappropriated secret with “reasonable particularity” early in the case may become the “case within the case.”

A recent example of a fight not to be the first party to identify source code in discovery is the New York state court case of MSCI, Inc. v. Jacob and Axioma, Inc. [pdf], where Supreme Court Judge Shirley Kornreich ruled the plaintiffs had to identify the components and sequencing of both the secret and the publicly known source code to the defendants, before discovery could proceed.

The plaintiffs claimed their former employee (who was not subject to a non-compete clause) had taken, used, and disclosed portions of source code to his new employer Axioma, who had embedded parts of that secret source code in its own financial market software.  The plaintiffs first characterized the entire source code “reference library” as a trade secret, even though parts of that code are in the public domain or licensed.  Their backup argument was that “certain details of the specific implementation” of even the publicly known code were trade secrets.  The plaintiffs proposed to only identify which parts of its source code were not trade secrets, and maintained this was a more practical and cost-efficient, yet legally sufficient way to meet its burden of trade secret identification.  An earlier preliminary ruling by the court affirmed that approach.

However, the issue came to a head when the plaintiffs sought full disclosure by the defendants of the Axioma source code library first.  The defendants argued that before the plaintiffs could obtain or inspect Axioma’s source code library, the plaintiffs should be made to reveal the precise parts of source code it claimed were trade secrets. 

Judge Kornreich was “persuaded that the law requires that a trade secret plaintiff identify trade secrets with reasonable particularity early in the case” (citing Xerox Corp. v. IBM Corp., 64 F.R.D. 367, 371 (S.D.N.Y. 1974) for the proposition that trade secret defendants should not be made to “guess at what they are” and that only after identification of a trade secret lost can a defendant “undertake a meaningful discovery program...”; and noting that California has codified this requirement at Cal. Code Civ. Proc. §2019.210).  Judge Kornreich reasoned that distinguishing between “general knowledge in the field” and trade secrets would allow the court to set the “parameters of discovery” and make the defendants “able to prepare their defense.”  The judge even observed that the plaintiffs, once privy to Axioma’s source code, “could tailor their theory of misappropriation” or even misappropriate Aximoa’s trade secrets. 

Before the plaintiffs can seek further discovery from the defendants, they must identify with reasonable particularity which of the “component parts or sequencing of their source code” are not publicly available, commonly used, or licensed from third parties.

This order has expensive and important ramifications in the case at issue, and generally, in source code misappropriation litigation.  A vital consideration in launching trade secret cases is the cost (dollar and competitive) of disclosure and identification of the secrets lost.