Score: 2

Prompting Away Stereotypes? Evaluating Bias in Text-to-Image Models for Occupations

Published: August 31, 2025 | arXiv ID: 2509.00849v1

By: Shaina Raza , Maximus Powers , Partha Pratim Saha and more

Potential Business Impact:

Makes AI art show different kinds of people.

Business Areas:
Text Analytics Data and Analytics, Software

Text-to-Image (TTI) models are powerful creative tools but risk amplifying harmful social biases. We frame representational societal bias assessment as an image curation and evaluation task and introduce a pilot benchmark of occupational portrayals spanning five socially salient roles (CEO, Nurse, Software Engineer, Teacher, Athlete). Using five state-of-the-art models: closed-source (DALLE 3, Gemini Imagen 4.0) and open-source (FLUX.1-dev, Stable Diffusion XL Turbo, Grok-2 Image), we compare neutral baseline prompts against fairness-aware controlled prompts designed to encourage demographic diversity. All outputs are annotated for gender (male, female) and race (Asian, Black, White), enabling structured distributional analysis. Results show that prompting can substantially shift demographic representations, but with highly model-specific effects: some systems diversify effectively, others overcorrect into unrealistic uniformity, and some show little responsiveness. These findings highlight both the promise and the limitations of prompting as a fairness intervention, underscoring the need for complementary model-level strategies. We release all code and data for transparency and reproducibility https://github.com/maximus-powers/img-gen-bias-analysis.

Country of Origin
🇨🇦 Canada

Repos / Data Links

Page Count
13 pages

Category
Computer Science:
Computation and Language