Twitter’s first bounty program for AI reveals these problems in its algorithm

A disabled automatic photo-cropping system by Twitter discriminates on the basis of colour, gender, weight and age, according to the results of an open competition to find algorithmic bias, media reports said.

The microblogging site had, in March, disabled the system as users noted that it favoured white and female faces while auto-cropping images. It then launched an algorithmic bug bounty competition offering prizes of up to $3,500 to analyse how the technology incorrectly handles photos, the Verge reported.

The top entry, contributed by Bogdan Kulynych, a graduate student in computer security at EPFL in Switzerland, showed that Twitter’s cropping algorithm favours faces that are “slim, young, of light or warm skin colour and smooth skin texture, and with stereotypically feminine facial traits”.