Skip to content

Wednesday, November 4, 2026No. 308

Anime database
Theme
anifr

Anime and manga news, announcements, trailers and industry reports with original sources.

Culture

VRChat: Despite 95 Percent Approval, No AI Labeling

21,595 people took part in the survey. Instead of AI labels, ratings and more preview images are meant to improve orientation.

Screenshot of a virtual three-dimensional environment in VRChat
VRChat is adjusting the rules for preview images of virtual worlds and how they are rated. · Image: VRChat / KAI-YOU

VRChat is not introducing mandatory labeling for AI-generated virtual worlds, even though 95 percent of respondents want some form of AI transparency label. The platform justifies its decision by pointing to inconsistent views on which working methods count as AI use, and to possible consequences for the community.

What the Survey Shows

As VRChat announced on October 29 Japan time, 21,595 people responded within one week. The results show broad support for more transparency, but hardly any agreement on the specific criteria. For eight proposed categories of possible AI use, 3,141 different combinations of labels emerged. No single combination reached more than ten percent of the answers.

33 percent say they want to deliberately avoid worlds with AI assistance. 37 percent of those who gave free-text answers said, in substance, that AI as a tool is acceptable as long as the human does the creative work. VRChat stresses that the figures were adjusted to the composition of its user base; for smaller groups, the results are less certain.

Why VRChat Is Not Introducing AI Labels

The operator fears false self-declarations, public accusations, and harassment of creators. The company also rejects a filter that hides AI content entirely. A voluntary label for worlds created entirely by humans, and information about the tools used, remain under discussion. No concrete introduction dates have been set for these yet.

Official VRChat graphic showing seven forms of AI use, from translation aids to fully generated virtual worlds
VRChat's official overview shows why a uniform definition of AI use is difficult. · Image: VRChat

Instead, VRChat already prohibits misleading preview images, regardless of the technique used. By 2026 or early 2027, users are to be able to rate virtual worlds positively or negatively. For the first half of 2027, multiple preview images per world are also planned.

Significance for Virtual Anime Projects

The rules also apply to worlds presenting anime and VTuber content. VRChat was present at the Tokyo Game Show 2026 with Hatsune Miku, Kizuna AI and other anime worlds, for example. The virtual exhibition on My Hero Academia and VR-Tsukuyomi for Chou Kaguya-hime! also show how important such spaces are for anime projects.

VRChat plans to provide further information on possible voluntary labeling before the end of March 2027. The operator's full statement explains both the results and the planned measures.

出典Sources

German original