Abstract
Criminal trials increasingly rely on computer programs to generate forensic evidence. But experts in the fi lds of computer science and forensic science often disagree over whether programs are sufficiently trustworthy to meet the legal admissibility standards for scientific evidence. When adjudicating between these disciplines, courts overwhelmingly side with forensic experts—even when considering technical, software-specific concerns. Usurping the intended application of the Daubert and Frye admissibility standards, courts blur the distinctions between scientific disciplines. Software experts are rendered unnecessary to establish the validity of software.
This Essay argues that courts habitually overestimate the reliability of software-generated evidence, and that courts do so because they grossly underestimate the specialized expertise involved in software validation. Instead, courts should treat the computer science and forensic science communities as discrete groups with distinct expertise over diffe ent aspects of forensic procedures. This would allow courts to develop a more robust application of current admissibility standards when considering increasingly interdisciplinary methods of generating evidence.
