Paper Overview
Field: NLP Authors: Ilana Nguyen, Harini Suresh, Thema Monroe-White Published: 2025-04-28 arXiv: 2504.19772
Summary
Large language models (LLMs) are increasingly used for text generation tasks, from everyday use to high-stakes enterprise and government applications, including simulated interviews with asylum seekers. While many works highlight the new potential applications of LLMs, there are risks of LLMs encoding and perpetuating harmful biases about non-dominant communities across the globe.
Key Findings
- The paper demonstrates persistent representational harms by national origin, including harmful stereotypes, erasure, and one-dimensional portrayals of Global Majority identities.
- Minoritized national identities are simultaneously underrepresented in power-neutral stories and overrepresented in subordinated character portrayals, which are over fifty times more likely to appear than dominant portrayals.
- Harms are amplified when US nationality cues are present in input prompts.
- These harms cannot be explained by sycophancy: US-centric bias persists even when American identity cues in prompts are replaced with non-US identities.
Original Abstract (excerpt)
> Large language models (LLMs) are increasingly used for text generation tasks from everyday use to high-stakes enterprise and government applications, including simulated interviews with asylum seekers. While many works highlight the new potential applications of LLMs, there are risks of LLMs encoding and perpetuating harmful biases about non-dominant communities across the globe. Our findings demonstrate the presence of persistent representational harms by national origin, including harmful stereotypes, erasure, and one-dimensional portrayals of Global Majority identities. Minoritized national identities are simultaneously underrepresented in power-neutral stories and overrepresented in subordinated character portrayals, which are over fifty times more likely to appear than dominant portra...
Read the full paper: https://arxiv.org/abs/2504.19772