Gender Bias in Text-to-Video Generation Models: A case study of Sora
This addresses bias in AI-generated content, which is a critical issue for fairness and representation in media, though it is incremental as it focuses on a specific model.
The study investigated gender bias in OpenAI's Sora text-to-video generation model, finding significant evidence that it disproportionately associates specific genders with stereotypical behaviors and professions, reflecting societal prejudices in its training data.
The advent of text-to-video generation models has revolutionized content creation as it produces high-quality videos from textual prompts. However, concerns regarding inherent biases in such models have prompted scrutiny, particularly regarding gender representation. Our study investigates the presence of gender bias in OpenAI's Sora, a state-of-the-art text-to-video generation model. We uncover significant evidence of bias by analyzing the generated videos from a diverse set of gender-neutral and stereotypical prompts. The results indicate that Sora disproportionately associates specific genders with stereotypical behaviors and professions, which reflects societal prejudices embedded in its training data.