Social Media: 'Forbidden' Words and Algorithm Censorship

Rumors circulate in the social media world about a secret list of 'forbidden words' that algorithms penalize, leading to the creation of the algospeak phenomenon, where users use coded language to avoid censorship.

Platforms like YouTube, Meta, and TikTok assure that there is no such list and that judgment is based on context and intent. However, content creators like Alex Perlman observe patterns suggesting otherwise.

Perlman reports that on TikTok he avoids mentioning YouTube, as the reach of his videos decreases. He also mentions problems with videos about Jeffrey Epstein, which were removed without clear explanation.

Platforms explain that algorithms aim to display content that interests the user, while community rules concern hate speech, violence, and misinformation. Nevertheless, behind-the-scenes interventions have been revealed, such as instructions to 'blur' content from specific users and the downgrading of pro-Palestinian content.

Users resort to 'algorithmic imagination,' creating myths about how algorithms work. A characteristic example is the use of the word 'festival' for demonstrations against ICE.

Companies deny the existence of a list of forbidden words, but content creators report patterns of punishment. Platforms filter content they deem dangerous, leading to self-censorship and a peculiar digital puritanism.