FaceApp Apologizes for ‘Hot’ Filter Preset Which Whitens User’s Skin

Viral face-tuning software FaceApp allows users to apply unconventional filters to photos of human faces, providing a gender changing, smiling, aging, de-aging filter and more. FaceApp utilizes a neural network to change the person’s facial characteristics, drawing from millions of photos in a database which the algorithm matches and edits into the photo almost instantly. However, one setting called the ‘hot’ filter, which supposedly beautifies a person’s appearance, removes wrinkles, brightens the eyes and apparently also whitens the skin. It does not sound that bad until you see the photo, which users on Twitter and other social media platforms have been posting:

Some users have called the app ‘racist’ for using white as the default hot preset, but Yaroslav Goncharov, the creator and CEO of FaceApp claims that it is actually an unfortunate side-effect of the underlying neural network caused by a training set bias, and not intended. He apologized if people are offended and have renamed the filter preset to ‘spark’ instead. FaceApp, uses deep learning technology to alter the photo automatically.

Self-beautification apps are not new, but FaceApp is one of the most advanced available. It is also not the first to draw the ire of the public for seemingly racist features. SnapChat came under fire last year for having a filter that turned faces into Asian caricatures and before that there was a Bob Marley feature which some took offense too and likened to a ‘digital blackface’.