Clearview AI finally takes part in a federal accuracy test

Source: Kashmir Hill
Affiliation: The New York Times

This article describes how the controversial facial recognition algorithm from Clearview AI was put to a third party test for the first time in 2021. The algorithm performed well on the test of matching two different photos of the same person. However, the algorithm was not tested on the one-to-many test, which finds a match for an unknown face amongst billions. The latter test is more in line with the algorithms current use cases in law enforcement.

Keywords: Computer Science , Facial Recognition , Tech Ethics