The Paradox of Source Code Secrecy

Abstract

Today, the government relies on machine learning and AI in predictive policing analysis, family court delinquency proceedings, parole decisions, and DNA and forensic science techniques, among other areas, producing a fundamental conflict between civil rights and automated decisionmaking. Ground zero for this conflict, I argue, has become the murky, messy intersection between software, trade secrecy, and public governance. In many cases of automated decisionmaking, algorithms – and the source code that informs them, are hidden from public view, even though they implicate core constitutional protections of due process, individualized justice and equal protection. However, because they are often protected as trade secrets, they can remain entirely free from public scrutiny.

This article argues that the constitutionally-inflected conflict that we now face is, in no small part, attributable to a core failure of our system of intellectual property to address, definitively, the boundaries of software protection and the implications for source code secrecy. In a world of privatized decisionmaking, the largely consistent move towards closed code in software sectors, has a number of deleterious results for the public, particularly in the age of algorithmic dominance. However, this Article argues that source code also carries a paradoxical character that is peculiar to software: the very substance of what is secluded often stems from the most public of origins, and often produces the most public of implications. And it is the failures of intellectual property law that has made this possible. First, as I show, courts have shifted the boundaries of protection for software under both copyright and patent law, further amplifying the attractiveness of trade secrecy. Second, the law has been willing to entertain an unique – and paradoxical-- overlap between copyright, patent, and trade secrecy, even though the three regimes have opposing public goals. Copyright and patent law are oriented towards disclosure, trade secrecy the opposite. While this overlap of protection in software seemed, at first glance, to be a good thing for innovation policy, it has proven deleterious for the larger public, particularly criminal defendants and lower income populations, who are now increasingly governed by an invisible hand that they can no longer investigate or question. But, as I argue, it may also be deleterious for other innovators, as well. The Article concludes with a brief discussion of ways to offer greater transparency through a "controlled disclosure regime," offering areas of reform in intellectual property, contract law, and discovery.

Keywords: Source code, Artificial Intelligence, Trade Secrets, Software