Exploring the Mechanism of Simulacral Identification Negotiation in AI-Generated Realistic Visuals: A Case Study of Genshin Impact Videos on Bilibili

With the growing use of artificial intelligence-generated content (AIGC) in visual production, AI-generated realistic visuals have emerged as a prominent creative practice within anime, comics, and games (ACG) communities. Judgements about whether AI-generated visuals “look like” original characters function not only as aesthetic evaluations, but also as sites for negotiating simulacral identification and cultural affiliation. This study examines a series of AI-generated realistic videos of Genshin Impact released on Bilibili, analysing viewer comments and danmaku (real-time bullet comments). Drawing on simulacra theory, the encoding/decoding model, and the concept of algorithmic visibility, this paper proposes a Three-Layer Model of Simulacral Identification Negotiation, comprising content generation, user cognition, and platform modulation. In the context of AIGC, the cultural meaning of visual simulacra emerges through the dynamic interaction of technical generation, user discourse, and platform mechanisms, which collectively reshape the expressive pathways of identity recognition and validation. The model holds cross-platform applicability, offering theoretical insight into how simulacral identification is constructed, circulated, and governed across global social media environments. Furthermore, the study offers insights into broader issues such as identity projection and cultural ethics, which increasingly shape the practice of visual re-simulacra in AIGC-driven media ecosystems.