Journal of Studies in Language, Culture and Society (JSLCS)
Volume 9, Numéro 2, Pages 176-187
2026-06-30

Selected Ai-powered Chatbots’ Patterns Of Identity Construction Of Three Major Ethnic Groups In Nigeria

Authors : Ugoala Bibian .

Abstract

The proliferation of chatbots in online interaction raises critical questions on how identity is constructed in human-machine discourse. This study aims to integrate insights from the Social Identity Theory and Cognitive Load Theory to examine how selected chatbots (Copilot, Gemini, and ChatGPT,) construct identity about three major ethnic groups in Nigeria. Three prompts in relation to identity construction about the three major ethnic groups in Nigeria were fed into each large language model, the responses they gave served as the data for the study. The specific objectives are: to unravel how chatbots through their responses in relation to (name, food, and dance) identify cultural traits of a particular ethnic group; to describe the pattern of “be” verb in chatbots’ responses to cultural identification of a particular ethnic group, and to investigate the pattern of chatbots’ responses to “wh” questions in relation to cultural identification of particular ethnic group. Chatbots generate responses from large data sets, however, this study argues that large language models’ responses to prompts does not truly represent the fluidity and naturalness which characterize human interactions. The study found that is some cases, the chatbots fail to properly distinguish the cultural identity of a particular region. They also in some cases in their responses, fail to respond in relation to the tense in which a prompt is structured. As chatbots gradually permeate all sectors, the study recommends that when seeking the true identity of an ethnic group, chatbots are not a reliable option, they lack the originality with which human interaction is known for.

Keywords

Artificial intelligence ; Chatbots ; ChatGPT ; Copilot ; Gemini