Table of Contents
Fetching ...

AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuances

Dhruv Agarwal, Mor Naaman, Aditya Vashistha

TL;DR

Findings show that Western-centric AI models homogenize writing toward Western norms, diminishing nuances that differentiate cultural expression.

Abstract

Large language models (LLMs) are being increasingly integrated into everyday products and services, such as coding tools and writing assistants. As these embedded AI applications are deployed globally, there is a growing concern that the AI models underlying these applications prioritize Western values. This paper investigates what happens when a Western-centric AI model provides writing suggestions to users from a different cultural background. We conducted a cross-cultural controlled experiment with 118 participants from India and the United States who completed culturally grounded writing tasks with and without AI suggestions. Our analysis reveals that AI provided greater efficiency gains for Americans compared to Indians. Moreover, AI suggestions led Indian participants to adopt Western writing styles, altering not just what is written but also how it is written. These findings show that Western-centric AI models homogenize writing toward Western norms, diminishing nuances that differentiate cultural expression.

AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuances

TL;DR

Findings show that Western-centric AI models homogenize writing toward Western norms, diminishing nuances that differentiate cultural expression.

Abstract

Large language models (LLMs) are being increasingly integrated into everyday products and services, such as coding tools and writing assistants. As these embedded AI applications are deployed globally, there is a growing concern that the AI models underlying these applications prioritize Western values. This paper investigates what happens when a Western-centric AI model provides writing suggestions to users from a different cultural background. We conducted a cross-cultural controlled experiment with 118 participants from India and the United States who completed culturally grounded writing tasks with and without AI suggestions. Our analysis reveals that AI provided greater efficiency gains for Americans compared to Indians. Moreover, AI suggestions led Indian participants to adopt Western writing styles, altering not just what is written but also how it is written. These findings show that Western-centric AI models homogenize writing toward Western norms, diminishing nuances that differentiate cultural expression.
Paper Structure (42 sections, 8 figures, 8 tables)

This paper contains 42 sections, 8 figures, 8 tables.

Figures (8)

  • Figure 1: Hofstede's Cultural Onion
  • Figure 2: The interface to complete the writing tasks in the AI condition. Suggestions were shown in grey next to the text written by the user and could be accepted by pressing the TAB key.
  • Figure 3: Various metrics of engagement with AI suggestions. Indian participants show a higher engagement with AI suggestions but also need to modify them more often.
  • Figure 4: Productivity measures. (a, b) Both cohorts experience productivity gains from writing with AI suggestions. (c) However, the productivity derived per suggestion is lower for Indian participants.
  • Figure 5: Within- and cross-culture similarity between Indian and American essays when written without and with AI. (a) The within-culture similarity is higher when AI is used, suggesting that AI homogenizes writing within both cohorts. (b) The cross-culture similarity is higher when AI is used, showing that AI homogenizes writing styles across cultures. (b) Cross-culture similarity scores are higher when AI suggestions are used.
  • ...and 3 more figures