Leaked insight into how Facebook censorship works

 
Hundreds of internal Facebook documents on sex, violence, and terrorism have been leaked. The documents give us an insight into how Facebook decides what we can read and view. The terminology used by Facebook is credible violence (aspirational conditional) versus credible violence (calls for action) Interestingly none of what was revealed explains Facebook’s history of censoring criticism of Islam and enforcing Islamic blasphemy laws.

Of the below three statements only one would be censored by Facebook can you guess which one?

    1. Someone shoot Trump
    2. Unless you stop bitching I will have to cut your tongue out.
    3. Kick a person with red hair
    4. To snap a bitches neck be sure to apply all your pressure to the middle of her throat

The answer is number one, “Someone shoot Trump.”

Internal company documents, leaked to The Guardian, detail the rules for allowing or rejecting content on violence, including live-streaming self-harm, racism, hate speech, terrorism and pornography.

In an investigation, the newspaper laid bare – for the first time – how the social media giant moderates content for its two billion users.

The investigation said the leaked documents included more than 100 internal training manuals, spreadsheets, and charts used to moderate users’ feeds.

…The report said the files showed the difficulties faced by human moderators confronted by an increasing volume of content such as revenge porn. As such, some of the moderators were concerned about the inconsistency and “peculiar nature” of some Facebook policies.

Under the violence policy, for example, the company attempts to draw distinctions between credible physical threats, verbal threats, and self-harm.

Blueprints have alarmed free speech advocates concerned about Facebook’s “de facto role as the world’s largest censor”, the report said.

screenshot-whaleoil

The investigation viewed documents relating to moderation policies within the last year and found remarks such as “To snap a bitch’s neck, make sure to apply all your pressure to the middle of her throat” were potentially permissible, as Facebook did not deem the comments as credible threats.

screenshot-whaleoil

…Violent videos, including self-harm, did not always have to deleted because they were marked as disturbing and could create awareness. Bullying videos, excluding sexual content, were permissible unless they were sadistic.

Livestreaming self-harm was also permissible because Facebook “did not want to censor or punish people in distress.”

“Generic” or “not credible” threats – and it was not clear how Facebook arrived at a definition of “not credible” – included “I hope someone kills you“.

…”We should say that violent language is most often not credible until specificity of language gives us a reasonable ground to accept that there is no longer simply an expression of emotion but a transition to a plot or design. From this perspective language such as ‘I’m going to kill you’ or ‘F*** off and die’ is not credible and is a violent expression of dislike and frustration.”

…”No matter where you draw the line there are always going to be some grey areas. For instance, the line between satire and humour and inappropriate content is sometimes very grey,”

…Videos of “non-sexual” child abuse were permissible, marked as disturbing content, and provided there was no sadism or celebration. Another slide said this could help identify and rescue a child…

 – Stuff


Do you want:

  • Ad-free access?
  • Access to our very popular daily crossword?
  • Access to Incite Politics magazine articles?
  • Access to Podcasts?
  • Access to Political Polls?

Our subscribers’ financial support is the reason why we have been able to offer our latest service; Audio blogs. 

Click Here  to support us and watch the number of services grow.

If you agree with me that’s nice but what I really want to achieve is to make you question the status quo. Look between the lines, do your own research. Do not be a passive observer in this game we call life.

39%