This article's reporting is false. They say "The researchers proposed a tweak so that hate speech against Black, Jewish, LGBTQ, Muslim, and multiracial people w

This article's reporting is false. They say "The researchers proposed a tweak so that hate speech against Black, Jewish, LGBTQ, Muslim, and multiracial people would be taken down through an automated system.", but the original Washington Post article they're summarizing (https://www.washingtonpost.com/technology/2021/11/21/facebook-algorithm-biased-race/ https://www.washingtonpost.com/technology/2021/11/21/faceboo...) says that the automated hate speech detection system already exists; the proposed tweak was to disable that system for hate speech against any groups other than the listed five, turning it back on only as an aspirational future goal when the algorithm "had been retrained and was on track". This is because the researchers weren't just aiming to quell hate speech in general - they wanted to change the fact that Facebook is (they concluded) "better at cracking down on comments that were harmful against White People".