Dataset of Multilingual GenAI News Analyses Across English, German, and Spanish

DOI

This dataset is part of the following publication:  Brinkmann, L.M. & Shafirova, L. (forthcoming). Exploring GenAI for multilingual news analysis and its pedagogical potential for the language classroom. In F. Palacios-Hidalgo & C. Lindade (Eds.), Equity, Equality, and Justice in AI-Enhanced Language Education. IGI Global.

The chapter examines the pedagogical potential of generative artificial intelligence (GenAI) for multilingual news analysis in language education. It situates GenAI within a multiliteracies framework, with a focus on critical multilingual language awareness and the need to address biases embedded in AI-generated content. The chapter explores how comparing GenAI outputs across languages can highlight differing perspectives and foster multiliteracies. Drawing on empirical analyses of news representations in English, German, and Spanish, it demonstrates how GenAI can be used for multilingual and multiliteracies purposes. The findings point to differences not only in presentation and content but also in what is foregrounded or omitted across languages. This underscores the value of using GenAI to analyse how meaning is shaped differently across linguistic contexts.

The data set is a comparison of a ChatGPT analysis, for which we used the BBC article “Why has Trump attacked Venezuela and taken Maduro?” by Vanessa Buschschlüter (January 3, 2026). The analysis was conducted through systematic comparisons of ChatGPT-generated responses across three language pairings: English/Spanish, English/German, and German/Spanish. The distribution of language pairings was determined by the researchers’ linguistic competencies, with only one author having proficiency in German. To present the data, we translated, using DeepL, examples into English for to enhance comprehension across languages. Each comparison followed a structured, step-by-step approach, examining the contextual dimensions of 1. Scenario / Setting; 1.1 Time; 1.2 Place; 1.3 Circumstances; 1.4 Type & means of communication; 2. Participants; 2.1 Identities; 2.2 Relationships; 3. Act(s) / Activity / Interaction; 3.1 Speech acts; 3.2 Communicative acts; 3.3 Social acts; 3.4 Political acts; 3.5 Social activity. 

Identifier
DOI https://doi.org/10.25592/uhhfdm.18689
Related Identifier IsPartOf https://doi.org/10.25592/uhhfdm.18688
Metadata Access https://www.fdr.uni-hamburg.de/oai2d?verb=GetRecord&metadataPrefix=oai_datacite&identifier=oai:fdr.uni-hamburg.de:18689
Provenance
Creator Brinkmann, Lisa Marie; Shafirova, Liudmila
Publisher Universität Hamburg
Publication Year 2026
Rights Creative Commons Attribution 4.0 International; Open Access; https://creativecommons.org/licenses/by/4.0/legalcode; info:eu-repo/semantics/openAccess
OpenAccess true
Representation
Resource Type Dataset
Discipline Other