by Brian Sites
16 Colum. Sci. & Tech. L. Rev. 36 (Published December 21, 2014)
Forensic machines and other such devices can create powerfully incriminating evidence, and some do so with little or no human aid. In a world where machines increasingly perform that task, what happens to the right to cross-examine the human witnesses who would have testified but for the machines? Through errors, bias, and even fabrication, machine operators, calibrators, and others involved in the forensic process still have the power to cause these tools to incriminate the wrong person. There is no shortage of evidence that this occurs. Many courts have concluded, however, that there is no right to cross-examine the operators of these machines. This Article analyzes that result.
The Article begins by discussing Confrontation Clause jurisprudence from across the nation, including machine-generated testimony cases and areas of Confrontation Clause jurisprudence that might be instructive in addressing machine-generated data, such as cases involving photographs, videos, interpreters, and dog handler testimony. The article then considers the strengths and weaknesses of several potential approaches to machine-generated testimony and concludes that multiple approaches are defensible, but the approach that best adheres to the purposes of the Confrontation Clause focuses on the degree and nature of the operator’s control over the machine. Under that model, there is no right to cross-examine the operators of many modern machines. Though that result is troubling, there are alternative models that might extend the lifespan of the Confrontation Clause. However, as machines become increasingly automated, the right to cross-examine their human assistants and progenitors will approach extinction.
For proper legal citation of this document, please cite to the following URL: http://www.stlr.org/cite.cgi?volume=16&article=2. The URL that currently appears in your browser’s location toolbar is incorrect. For more information on Bluebook citation of internet sources, click here.