Classification accuracy of the event-related potentials-based Brain Fingerprinting and its robustness to direct-suppression and thought-substitution countermeasures

Abstract

Research on the accuracy of Brain Fingerprinting (BFP) has produced mixed outcomes: some report 99.9% and others report lower. Furthermore, no studies have measured the susceptibility of BFP to countermeasures. In Experiment-1, we report the accurate classification of 15 of the 16 subjects, tested on their own real-life autobiographical incidents; and 14 of the 15 other subjects, tested on another subject's real-life autobiographical incidents. In Experiment-2, 16 subjects of Experiment-1, who were tested on their own real-life incidents, participated in the BFP test again, but this time employing either direct-suppression or thought-substitution (n = 8 each) countermeasures. We report that neither direct-suppression nor thought-substitution was effective at concealing information that BFP was designed to reveal. We assert that BFP is a highly accurate, albeit not perfect, concealed-knowledge detection technology and that it is resistant to memory suppression and thought substitution countermeasures in the context of autobiographical incidents.