PulseAugur
EN
LIVE 20:37:56

LLMs show language and demographic biases in new research

By PulseAugur Editorial · [4 sources] ·

New research indicates that multilingual large language models exhibit significant language biases, often favoring certain languages over others when presented with conflicting information. Studies also reveal disparities in gender, racial, and age representation within LLMs, with debiasing efforts sometimes creating new fairness trade-offs. These models frequently deviate from real-world demographic data in occupational and crime scenarios, and their stereotyping can be amplified across different languages. AI

IMPACT Highlights critical biases in LLMs that could affect fairness and reliability in real-world applications, necessitating improved mitigation strategies.

RANK_REASON Multiple academic papers published on arXiv detailing LLM bias evaluations.

Read on arXiv cs.AI →

AI-generated summary · Google Gemini · from 4 sources. How we write summaries →

LLMs show language and demographic biases in new research

COVERAGE [4]

  1. arXiv cs.CL TIER_1 English(EN) · Robert \"Ostling, Murathan Kurfal{\i} ·

    Language Bias under Conflicting Information in Multilingual LLMs

    arXiv:2604.07123v2 Announce Type: replace Abstract: Large Language Models (LLMs) have been shown to contain biases in the process of integrating conflicting information when answering questions. Here we ask whether such biases also exist with respect to which language is used for…

  2. arXiv cs.AI TIER_1 English(EN) · Ikhlasul Akmal Hanif, Muhammad Falensi Azmi, Filbert Aurelian Tjiaranata, Eryawan Presma Yulianrifat, Fajri Koto ·

    IndoBias: A Dual Track Culturally Grounded Benchmark for LLMs Bias Evaluation in Indonesian Languages

    arXiv:2606.01260v1 Announce Type: cross Abstract: Despite being home to more than 1300 ethnic groups and 700 indigenous languages, bias in Large Language Models has not been fully studied in Indonesia, thus leaving a critical gap in evaluating representational fairness and locali…

  3. arXiv cs.AI TIER_1 English(EN) · Vishal Mirza, Rahul Kulkarni, Aakanksha Jadhav ·

    LLM Bias Evaluation: Gender, Racial, and Age Disparities in Occupational and Crime Scenarios

    arXiv:2409.14583v4 Announce Type: replace Abstract: LLM bias evaluation is critical as large language models (LLMs) increasingly influence high-stakes decisions. This paper provides a comprehensive assessment of gender, racial, and age disparities in leading LLMs, revealing that …

  4. arXiv cs.CL TIER_1 English(EN) · Jiwoo Choi, Seonwoo Ahn, Tongxin Zhang, Seohyon Jung ·

    Anchoring LLM Gender Bias to Human Baselines: A Cross-Lingual Audit

    arXiv:2605.30804v1 Announce Type: new Abstract: We audit six large language models (LLMs) for gender stereotyping across English, Korean, Chinese, and Japanese. Three were developed primarily for English-language use (Claude, GPT, Gemini) and three for East Asian use (DeepSeek, S…

RELATED ENTITIES

RELATED TOPICS