Fluent but Foreign: Even Regional LLMs Lack Cultural Alignment

TL;DR

This paper evaluates whether regional LLMs reflect local cultural values and practices, finding that they often do not, and highlights the need for culturally grounded data and evaluation methods.

Contribution

It introduces a methodology for culturally grounded evaluation of LLMs and demonstrates that current regional models lack true cultural alignment due to data scarcity.

Findings

Indic models do not better align with Indian norms than global models

Both global and Indic models produce Westernized or exoticized writing suggestions

Prompting and fine-tuning often degrade cultural alignment

Abstract

Large language models (LLMs) are used worldwide, yet exhibit Western cultural tendencies. Many countries are now building ``regional'' or ``sovereign'' LLMs, but it remains unclear whether they reflect local values and practices or merely speak local languages. Using India as a case study, we evaluate six Indic and six global LLMs on two dimensions -- values and practices -- grounded in nationally representative surveys and community-sourced QA datasets. Across tasks, Indic models do not align better with Indian norms than global models; in fact, a U.S. respondent is a closer proxy for Indian values than any Indic model. We further run a user study with 115 Indian users and find that writing suggestions from both global and Indic LLMs introduce Westernized or exoticized writing. Prompting and regional fine-tuning fail to recover alignment and can even degrade existing knowledge. We attribute this to scarce culturally grounded data, especially for pretraining. We position cultural evaluation as a first-class requirement alongside multilingual benchmarks and offer a reusable, community-grounded methodology. We call for native, community-authored corpora and thickxwide evaluations to build truly sovereign LLMs.