Facebook Explains How It Defines Hate Speech And Why So Much Is Able To Slip Through The Cracks

As part of its mission to protect its users, Facebook pledged to answer the hard questions about its operations and it is making good on that promise by detailing its policy on hate speech.

Hate speech is difficult to define and the very topic is controversial. Some advocates of absolute free speech argue that no speech should be restricted, but many organizations and governments have policies restricting hate speech. Facebook is in that latter category and its definition of hate speech is pretty clear-cut, at least on the surface.

Facebook's Definition Of Hate Speech

"Our current definition of hate speech is anything that directly attacks people based on what are known as their 'protected characteristics' — race, ethnicity, national origin, religious affiliation, sexual orientation, sex, gender, gender identity, or serious disability or disease," Facebook said in a statement.

However, even with this definition, things aren't always easy. As Facebook's blog points out some people might consider an offensive religious joke an example of hate speech so the waters are a bit murky. Another example is the fact that the use of certain racial slurs would normally constitute hate speech, but what if the poster in question is merely quoting a song?

Facebook's Policy On Hate Speech

Facebook's policy is to remove hate speech whenever it is encountered on the site and, as it turns out, there's a lot there. Over the past two months, the site deleted 66,000 reported cases of hate speech per week, which averages to roughly 288,000 deletions a month.

Unfortunately, as Facebook readily admits, it occasionally gets things wrong. One of the most notable examples was the one that occurred involving the African-American activist Shaun King. King had posted a piece of hate mail that he had received which contained numerous racial slurs. Facebook removed the post as a violation of its policy on hate speech not realizing that King had shared the letter to make a point about some of the abuse he receives. Facebook did correct the issue once it was brought to its attention.

The Use Of Artificial Intelligence To Counter Hate Speech

Facebook did note that it was attempting to use AI in order to help combat hate speech. For example, the blog noted that it is creating AI that will hide posts that contain offensive language. However, the AI is far from perfect and runs into the issue of context. The issue of King's post is a prime example. An AI that is trained to block posts containing racial slurs might have blocked King's post without taking into account the context as to why it was posted.

ⓒ 2018 All rights reserved. Do not reproduce without permission.
Real Time Analytics