Bilibili is a Chinese online video-sharing platform that has evolved into a cultural phenomenon among the country's youth. Initially launched in 2009 as a niche community for enthusiasts of anime, comics, and games (ACG), it has since expanded to encompass a wide array of content, including music, dance, science and technology, movies, drama, fashion, and daily life videos.
One of Bilibili's distinctive features is its "danmu" (弹幕) system, which allows real-time, overlaid comments to appear on videos, creating a highly interactive and communal viewing experience. This feature fosters a sense of shared engagement, as users can see and respond to each other's reactions both synchronously and asynchronously.
In summary, Bilibili serves as a multifaceted platform that blends user-generated content, interactive features, and a strong community focus, making it a central hub for young Chinese netizens seeking diverse and engaging online experiences.
Platform Support and Regulation
Bilibili has emerged as a vibrant hub for AI-generated singing content, nurturing a dynamic community of creators and fans. This trend garnered nationwide attention with the viral success of AI-generated songs featuring the voice of Singaporean singer Stefanie Sun, surpassing the platform’s usual niche and even being featured on national news programs. The widespread popularity of these AI renditions sparked public discourse around authenticity and authorship. Sun herself also acknowledged the phenomenon, voicing both fascination and concern over the implications of AI on creative industries.
In response to the growing prevalence of such content—and the accompanying concerns surrounding copyright and authenticity—Bilibili introduced a labeling system to clearly identify AI-generated videos. This initiative aims to promote transparency and preserve the integrity of content on the platform. Unlike YouTube, which largely restricts AI-singing due to copyright issues, Bilibili actively encourages this creative practice, fostering the development and popularity of a distinct subgenre within its digital music culture.
A Sharing Community
Bilibili’s fan-driven culture encourages a strong tradition of sharing, often expressed through the meme Weiai fadian (为爱发电, “generate electricity from love”). In this regard, fan works are “labors of love,” and many creators on Bilibili produce AI-singing content not for profit, but out of passion. For instance, many users have uploaded tutorial videos (the one on the left is an example) that teach others how to use Sovits to create their own AI voice models. In the “What is Sovits” section, you’ll also find tutorials in multiple languages, making the process more accessible to a global audience.
As the homepage shows, many users openly share their AI Dongxuelian voice models, making it easier for others to create, remix, and build on existing works. This informal “Creative Commons” allows the community to thrive through collaboration and mutual appreciation.
From Guichu (鬼畜) to AI Singing: A Remix Lineage
The rise of AI-singing on Bilibili doesn’t emerge in isolation — it builds on a vibrant foundation of remix culture on Bilibili, particularly the genre of Guichu (鬼畜). Known for its humorous edits, pitch-shifted vocal loops, and rhythmic cut-ups of political figures, celebrities, or viral personalities, Guichu represents a form of creative user-driven remix (rooted in Chinese online culture). What links Guichu and AI-singing is not just aesthetic play, but a shared ethos of “re-authoring the voice” — taking existing vocal material and reanimating it through technology and fan labor.
In this sense, AI-singing based on Sovits can be seen as a technological evolution of Guichu’s practices, shifting from manual auto-tuning to full-scale vocal synthesis. The fandom infrastructure that supported Guichu — open-source tools, tutorial sharing, comment culture — now sustains the AI-singing community. Moreover, both genres destabilize the idea of authorship, challenge notions of vocal “authenticity,” and invite collective participation in voice-based creativity. In this regard, AI-singing continues and amplifies Bilibili’s remix legacy and aesthetics.
A news article introducing Guichu in English: https://www.sixthtone.com/news/1003118
The Guichu Section on Bilibili (Separated from Music section)