Several social media platforms are working to minimize ‘hate speech’ on their platforms, to create a healthier environment–but some have criticized these plans as censorship, promoting certain political ends.
Facebook has been known to collude with foreign governments to police hate speech. If certain posts aren’t removed in a timely manner, governments have threatened to fine the social media company for their inaction.
A Swedish man landed in hot water after failing to remove ‘hateful’ comments made by other people in a group he was running.
While AI is deployed to filter social media for hate speech committed by potential ‘right-wing extremists’ borne out of fears of violent attacks, there has been a politically incorrect twist to the narrative: black people crowned as worst offenders for social media ‘hate speech.’
Two studies confirm African-Americans as leaders of online ‘hate.’
One study found that black Twitter users were one-and-a-half times likelier to have tweets flagged as ‘hateful.’ Vox reports that speech written in “African American English” were flagged 2.2 times higher that traditional US English.
Another study claims to have found widespread spread bias against African-American speech in general.
One of the various complaints over the findings has been the AI bots’ inability to determine context–as some words such as the “n-word” or “queer” aren’t offensive if used by certain people.
Activists have claimed that black people are unfairly singled out for their use of language, and feel as if they’re being harshly targeted.
Some believe that the faultiness of the software merely reinforces racial biases.