The 8088 The 8088 ← All news
arXiv cs.CL AI Research Apr 21

IYKYK (But AI Doesn't): Automated Content Moderation Does Not Capture Communities' Heterogeneous Attitudes Towards Reclaimed Language

★★★★★ significance 3/5

This research examines how automated content moderation tools fail to distinguish between hateful language and the use of reclaimed slurs within marginalized communities. The study highlights significant disagreement among human annotators regarding slur usage, which leads to the suppression of voices in LGBTQIA+, Black, and women's communities.

Why it matters Algorithmic moderation failures risk systemic silencing of marginalized communities by misidentifying reclaimed language as prohibited speech.
Read the original at arXiv cs.CL

Tags

#content moderation #reclaimed language #bias #nlp #social impact

Related coverage