We don't have AI content moderation systems that catch harmful content without also flagging large amounts of legitimate content.
open
Global / Unspecified, Global
Automated moderation systems still struggle to tell the difference between genuinely harmful content and similar-looking legitimate content, like educational or satirical material. This tradeoff between over- and under-moderation remains unresolved at scale.
Citation ID: WS00697
Title: We don't have AI content moderation systems that catch harmful content without also flagging large amounts of legitimate content.
URL: https://www.worldsolve.org/index.php?api=problem&id=697