The Democratization and Dilemmas of AI Text-to-Video Generation: A Comprehensive Survey Analysis

Abstract

AI text-to-video systems have moved rapidly from research prototypes to widely accessible creative tools. This paper surveys the current landscape through two lenses: (1) a scoping review of state-of-the-art models and governance mechanisms, and (2) an empirical, mixed-methods survey of creators, educators, and media professionals (N = 412) on access, use cases, perceived benefits, and risks. We find broad evidence of “democratization”—reduced cost and skill barriers enabling newcomers to produce compelling motion content—alongside unresolved dilemmas around provenance, copyright, deepfakes, bias, and compute intensity. We synthesize technical and policy directions to strengthen safety and accountability while preserving creative opportunity.