Drexel University

College of Computing and Informatics

Analysis of Facial Processing Technology

Analyze commercial facial processing technology, specifically Amazon Rekognition, to identify ethical vulnerabilities in the development of legal regulation and its integration into law enforcement.
The disequilibrium created by rapid technological innovation, particularly within artificial intelligence, combined with the legal system's inability to keep pace, results in the need to analyze its impact from an ethical standpoint. Facial processing technology (“FPT”), including facial detection, analysis, and recognition, is at the forefront of criminal investigations. Prior research into FPT implied a possible racial and gender bias, potentially exacerbating a prejudice already present in law enforcement, thus affecting individuals' civil liberties.
This project analyzes commercial facial processing software to explore patterns of potential bias to foresee possible ethical vulnerabilities regarding law enforcement’s use of FPT. Despite sample size restrictions, evidence of inaccuracy trends was found, giving strength to the potential validity of these claims.
FPT is examined from a utilitarian and deontological perspective, evaluating ethical concerns according to Marx’s Seven Stages of Issue Development and Fink’s Crisis Model. It was determined that without legislation regulating the use of FPT in criminal investigations, society will advance to the crisis stage. Based on this, a legislative approach was proposed. 
Logo Reference: https://welpmagazine.com/wp-content/uploads/2020/09/facial-recognition-ibm-1440x920-1-758x484.jpg

Team Members