31 Pages Posted: 17 Apr 2016
Date Written: April 13, 2016
The article deals with two legal issues posed by the growing trend in the United States to automate forensic analyses.
Since World War II, we have had alarming insights into the unreliability of both eyewitness testimony and confession evidence. Those insights have prompted the criminal justice system to place greater reliance on forensic evidence. In one Rand study, the researchers found that expert testimony was presented at 86% of the trials examined. This shift to greater use of expert testimony has placed growing demands on crime laboratories. For example, the backlog of unanalyzed DNA samples has become such an acute problem that Congress was impelled to enact the DNA Backlog Elimination Act to provide funding to reduce the backlog of untested rape kits.
In both the public and private sectors, the typical response to the development of a backlog is technological automation. That has certainly held true for forensic analysis. There is now widespread automation in such areas as fingerprint examination, breath testing, and DNA analysis. The argument runs that automation holds the promise of both enhancing efficiency and improving the accuracy of the analyses proffered in court.
That promise turns on the accuracy of the source code controlling the software governing the automated techniques. The source code embeds the instructions determining which tasks the program performs, how the program performs them, and the order in which it performs the tasks. The validity of a program’s source code is the most fundamental guarantee of a software program’s reliability. Defense counsel have sometimes challenged the software for automated forensic techniques. Early in this century, the defense counsel attacked the software controlling automated infrared breath testing devices. Today they are challenging the software for the TrueAllele program analyzing mixed DNA samples. Those waves of cases have posed two issues: (1) whether the prosecution can lay a sufficient foundation for evidence based on an automated technique without presenting testimony about the computer source code; and (2) whether the defense has any discovery right to access to the code. Almost all the courts have answered the first question in the affirmative and the second question in the negative. In responding to the second question, the courts have reasoned that the existence of validation studies for the technique eliminates any need to scrutinize the source code and that in any event, manufacturers have an evidentiary privilege protecting the code as a trade secret. The purpose of this short article is to critically evaluate the judicial response to both questions.
On the one hand, the article argues that the courts have correctly answered the first question. More specifically, the prosecution may lay an adequate foundation by presenting testimony describing validation studies for the automated technique even if the testimony does not touch on the source code. On the other hand, the article contends that in some cases, the courts ought to accord the defense a pretrial discovery limit. The article explains the limited utility of validation studies and notes that the evidentiary privilege for trade secrets is a qualified one that can be surmounted when the party seeking discovery has a significant need for the information. The article proposes a procedure that judges can employ to resolve the tension between the defendant’s need for access to the source code and the manufacturer’s legitimate interest in safeguarding its valuable proprietary information.
Suggested Citation: Suggested Citation
Imwinkelried, Edward J., Computer Source Code: A Source of the Growing Controversy Over the Reliability of Automated Forensic Techniques (April 13, 2016). DePaul Law Review, Forthcoming; UC Davis Legal Studies Research Paper No. 487. Available at SSRN: https://ssrn.com/abstract=2764593