Common sense or censorship: How algorithmic moderators and message type influence perceptions of online content deletion

Hateful content online is a concern for social media platforms, policymakers, and the public. This has led high-profile such as Facebook, to adopt algorithmic content-moderation systems; however, impact of moderation on user perceptions unclear. We experimentally test extent which type being removed (profanity vs hate speech) explanation given its removal (no link community guidelines specific explanation) influence human moderators. Our preregistered study encompasses representative samples ( N = 2870) from United States, Netherlands, Portugal. Contrary expectations, our findings suggest that perceived more transparent than human, especially when no removal. In addition, sending users further information deletion negative effects outcome fairness trust.