Facebook officially responds to the Wall Street Journal’s allegations regarding the poor efficiency of the artificial intelligence algorithm as for the removal of content related to hate speech. The issue is very current for the social network that is trying to put into practice increasingly effective countermeasures to counteract content of this type.
Hate speech posts would be in sharp decline on Facebook
According to a recent post published by Guy Rosen, Facebook’s VP of Integrity in these hours, hate speech content is in sharp decline on social media. In the last three quarters, in fact, there has been one reduction of about 50% of the incidence of such contents on the total. Currently, the data confirms that hate speech content accounts for 0.05% of the total posts viewed on Facebook.
It should be noted, underlines the social network, that the penalized content (removed from the News Feed and suggestions) and not just those actually removed from the platform. Facebook also adds that less than 25% of deleted content is automatically detected by the AI. Most of the removed content is reported by users of the platform.
Maximum transparency in the fight against hatred
In the post on the official blog, Facebook underlines the desire to completely eliminate hate speech posts from the social network. The countermeasures put in place to counter such practices will be characterized by total transparency.