Popularity as Algorithmic Dependency: How Bulgarian Social Media Users Perceive AI Influencers

Abstract

This study investigates the influence of platform architecture on the visibility, social legitimacy, and user engagement of AI influencers across TikTok, Facebook, and Instagram. By comparing platform metrics, algorithmic distribution, and user engagement patterns, it demonstrates how identical synthetic personas attain varying social statuses depending on the specific communication environment.

The findings indicate that the popularity of virtual influencers is shaped primarily by the cultural relevance of the characters and the logic of algorithmic distribution, rather than by the extent of visual human-like realism. TikTok enables rapid organic dissemination and high engagement through interest graph mechanisms, whereas Facebook fosters more conservative interaction patterns. Conversely, Instagram is less conducive to the success of deliberately unpolished AI personas. This study constitutes one of the first investigations in the Bulgarian context to analyze how platform logic influences public perceptions of synthetic identities.