Easily Accessible Text-to-Image Generation Amplifies Demographic Stereotypes at Large Scale

TL;DR

This paper reveals that accessible text-to-image models tend to amplify harmful stereotypes across various prompts, persisting despite mitigation efforts, thus raising concerns about their societal impact.

Contribution

It provides a comprehensive analysis of how current text-to-image models perpetuate stereotypes, highlighting their widespread presence and resistance to mitigation strategies.

Findings

Stereotypes appear even with neutral prompts

Mitigation strategies do not eliminate stereotypes

Models reinforce racial, gender, and cultural biases

Abstract

Machine learning models that convert user-written text descriptions into images are now widely available online and used by millions of users to generate millions of images a day. We investigate the potential for these models to amplify dangerous and complex stereotypes. We find a broad range of ordinary prompts produce stereotypes, including prompts simply mentioning traits, descriptors, occupations, or objects. For example, we find cases of prompting for basic traits or social roles resulting in images reinforcing whiteness as ideal, prompting for occupations resulting in amplification of racial and gender disparities, and prompting for objects resulting in reification of American norms. Stereotypes are present regardless of whether prompts explicitly mention identity and demographic language or avoid such language. Moreover, stereotypes persist despite mitigation strategies; neither user attempts to counter stereotypes by requesting images with specific counter-stereotypes nor institutional attempts to add system ``guardrails'' have prevented the perpetuation of stereotypes. Our analysis justifies concerns regarding the impacts of today's models, presenting striking exemplars, and connecting these findings with deep insights into harms drawn from social scientific and humanist disciplines. This work contributes to the effort to shed light on the uniquely complex biases in language-vision models and demonstrates the ways that the mass deployment of text-to-image generation models results in mass dissemination of stereotypes and resulting harms.