VRChat rejects AI labels over accuracy and harassment concerns

VRChat will not introduce dedicated labels for AI-generated content for now, despite a survey showing that 95 per cent of respondents wanted some form of disclosure. Users also will not receive a search filter to exclude AI-generated creations from the virtual reality social platform.

The company said a disclosure system would depend on creators accurately describing their work, leaving room for misleading claims about content being made entirely by hand. It also warned that labels could encourage users to investigate suspected AI use and harass creators, including those who do not use AI tools.

VRChat continues to consider a Human-authored label and a separate Tools used field, which could let creators describe their production methods and the extent of human involvement. Neither feature is confirmed, and the company has announced no implementation timeline.

Instead, VRChat plans to prioritize changes with broader community agreement, including more accurate thumbnails, better discovery tools and a stronger emphasis on quality when presenting content. Its current approach leaves disclosure dependent on information individual creators choose to provide, rather than dedicated platform tags.