Software that predicts whether you look like, and so will be, a criminal

Harrisburg University proudly announces, in a press release:

HU facial recognition software predicts criminality

A group of Harrisburg University professors and a Ph.D. student have developed automated computer facial recognition software capable of predicting whether someone is likely going to be a criminal.

With 80 percent accuracy and with no racial bias, the software can predict if someone is a criminal based solely on a picture of their face. The software is intended to help law enforcement prevent crime.

Ph.D. student and NYPD veteran Jonathan W. Korn, Prof. Nathaniel J.S. Ashby, and Prof. Roozbeh Sadeghian titled their research “A Deep Neural Network Model to Predict Criminality Using Image Processing.” …

The press release does not specify whether the three co-researchers applied the software to photographs of themselves. Here are photos of the co-researchers:

(Thanks to Michael Petrov for bringing this to our attention.)

UPDATE (May 11, 2020): Michael Beaver alerted us to two other studies that go adventuring in similar territory:

Automated Inference on Criminality using Face Images” [2016]

Criminal tendency detection from facial images and the gender bias effect” [2020]