The use of technology in human expert domains: challenges and risks arising from the use of automated fingerprint identification systems in forensic science

Abstract

Cognitive technologies have increased in sophistication and use, to the point of interactively collaborating and distributing cognition between technology and humans. The use of Automated Fingerprint Identification Systems (AFIS), a computerized database of fingerprints, by latent fingerprint experts is a par-excellence illustration of such a partnership in forensic investigations. However, the deployment and use of cognitive technology is not a simple matter. If the technology is going to be used to its maximum potential, we must first understand the implications and consequences of using it, and make whatever adaptations are necessary both to the technology and to the way humans work with it. As we demonstrate with AFIS, latent fingerprint identification has been transformed by the technology, but the strategies used by humans who work with this technology have not been modified and adequately adjusted in response to the reality of these transformations. For example, the chances that an AFIS search will produce prints with incidental similarities –i.e., that highly similar, look-alike, prints from different sources will result from an AFIS search – has not been sufficiently investigated or explored. This risk, as well as others, may mean that the use of AFIS introduces new concerns into the process of latent fingerprint identification, some of which may even increase the chances of making erroneous identifications. Only by appropriate and explicit adaptation to the new potential and the new challenges posed by the new technology will AFIS and other cognitive technologies produce efficient and effective partnerships.