Facebook apologizes for calling Black men "primates", in a featured YouTube video.
According to The New York Times, it was an "unacceptable mistake."
This label was used in connection to the clip "White man calls police on black men at marina"
10 Things in Tech: Keep up-to-date with the latest tech trends and innovations. Loading... Click to sign up for Insider marketing emails. You also agree to receive partner offers.
Facebook issued an apology Friday after its AI software called Black men "primates", in a video posted on the social media site.
The New York Times reported the story first. Facebook spokesperson said that the error was "clearly unacceptable" and that the recommendation software had been disabled.
The spokesperson stated that they had disabled all topic recommendation features as soon as they realized it was happening to allow them to investigate and prevent future incidents.
Facebook issued a statement to the publication saying: "We apologize for anyone who may have been offended by these recommendations."
The video was posted by The Daily Mail on June 27, 2020. The offensive language was related to the video. The video was called "white man calls police on black men at marina" and featured Black men engaging in dispute with civilians and white police officers.
According to The New York Times, Facebook users who viewed the video received an automatic prompt asking them if they'd like to "keep watching videos about primates"
Insider asked Facebook for comment but did not receive a response immediately.
Darci Groves was a former content manager at Facebook. The New York Times reported that a friend screen retrieved the recommendation and sent them to her. It was shared on Twitter by Darci Groves, who also posted it to a forum for product feedback from current and former Facebook employees.
Facebook Watch's product manager responded that the company was investigating the root cause.
Groves stated that the prompt was "horrifying" and "egregious."
In the past, technology companies like Google and Twitter have been accused of biases in their artificial intelligence software.
Twitter investigated the possibility that its automatic tool for selecting which parts of a photo to show in tweets might have racially biased bias against Black people last year, Insider's Isobel Hamilton reported. The company announced that it would reward anyone who could solve the problem in a new competition with a $3500 reward.
Google's algorithm in 2015 reportedly used the word "gorilla" to tag two Black faces. In a statement to The Wall Street Journal, the company stated that it was "genuinely sorry" for what happened.