Sony PlayStation uses machine learning classifiers for content moderation on its platform and has worked to address bias in these systems, specifically the problem of 'celebratory identity speech' being incorrectly flagged as discriminatory content.

A machine learning engineer at Sony PlayStation presented research at the Trust and Safety Research Conference at Stanford on the topic of bias mitigation in content moderation classifiers. The specific issue addressed was 'celebratory identity speech' — text that mentions protected groups in a positive context but is incorrectly scored as likely discriminatory by classifiers. This confirms the existence of ML-based content moderation classifiers at PlayStation, though no further details about their scope, deployment scale, or the full range of content they moderate were disclosed in available primary sources.