{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T06:27:51Z","timestamp":1786343271595,"version":"build-2736575974"},"reference-count":42,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T00:00:00Z","timestamp":1786320000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Large language models now handle a share of cross-lingual political translation, but no study has directly measured whether they follow stable normative preferences when doing so.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>We target Chinese-to-English translation of Chinese political and policy terms embedded in authentic official discourse. Mapping 52 publications across translation theory, LLM empirics, and AI alignment yields a five-dimensional Translation Norm Orientations (TNO) framework: Faithfulness (FN), Fluency (FLN), Cultural Adaptation (CAN), Ethical Hedging (EHN), and Foreignization Retention (FRN). Eight mainstream Chinese and American LLMs complete a Forced-Choice Translation Ranking Task (FCTRT) on 20 ideologically sensitive political terms, ranked across three repeated trials. Kendall\u2019s W assesses ranking stability, Mann\u2013Whitney U tests with BH-FDR correction test between-group differences, and hierarchical clustering identifies latent normative prototypes.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>\n                      All eight models rank FN first; CAN, EHN, and FRN settle into a shared low-priority band beneath it. None of the five dimensions shows a statistically significant CN\u2013US split. FN alone carries a medium effect size (Cliff\u2019s\n                      <jats:italic>\u03b4<\/jats:italic>\n                      \u202f=\u202f0.50), which points to a difference in how strongly, not which way, the two groups lean. Cluster membership cuts across national lines rather than following it: four transnational profiles emerge, separated by foreignization tolerance rather than developer geography.\n                    <\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>These results tie translation norm theory to AI evaluation and supply a benchmark other researchers can reuse to examine LLM-mediated political communication.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/frai.2026.1869972","type":"journal-article","created":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T05:30:19Z","timestamp":1786339819000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Faithful or evasive? An empirical study on translation norm preferences of Chinese and American LLMs in Chinese official political and policy discourse"],"prefix":"10.3389","volume":"9","author":[{"given":"Xinyu","family":"Zhu","sequence":"first","affiliation":[{"name":"School of Foreign Languages, Southeast University","place":["Nanjing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yu","family":"Yin","sequence":"additional","affiliation":[{"name":"School of Foreign Languages, Southeast University","place":["Nanjing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,10]]},"reference":[{"key":"ref1","first-page":"15763","author":"Adilazuarda","year":"2024"},{"key":"ref2","volume-title":"Translation and Conflict: A Narrative Account","author":"Baker","year":""},{"key":"ref3","volume-title":"In Other Words: A Coursebook on Translation","author":"Baker","year":""},{"key":"ref4","volume-title":"La traduction et la lettre, ou, L\u2019auberge du lointain","author":"Berman","year":"1999"},{"key":"ref5","first-page":"17140","author":"Chen","year":"2024"},{"key":"ref6","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1075\/target.5.1.02che","article-title":"From \u2018is\u2019 to \u2018ought\u2019: laws, norms and strategies in translation studies","volume":"5","author":"Chesterman","year":"1993","journal-title":"Target"},{"key":"ref7","doi-asserted-by":"crossref","DOI":"10.1075\/btl.123","volume-title":"Memes of Translation: The Spread of Ideas in Translation Theory","author":"Chesterman","year":"2016"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"37","DOI":"10.1177\/001316446002000104","article-title":"A coefficient of agreement for nominal scales","volume":"20","author":"Cohen","year":"1960","journal-title":"Educ. Psychol. Meas."},{"key":"ref9","doi-asserted-by":"publisher","first-page":"28","DOI":"10.1007\/s10676-025-09837-2","article-title":"Helpful, harmless, honest? Sociotechnical limits of AI alignment and safety through reinforcement learning from human feedback","volume":"27","author":"Dahlgren Lindstr\u00f6m","year":"2025","journal-title":"Ethics Inf. Technol."},{"key":"ref10","first-page":"46","author":"Freitag","year":"2022"},{"key":"ref11","doi-asserted-by":"crossref","first-page":"14788","DOI":"10.18653\/v1\/2024.findings-emnlp.868","volume-title":"Findings of the Association for Computational Linguistics: EMNLP 2024","author":"Gete","year":"2024"},{"key":"ref12","first-page":"112282","article-title":"Validating LLM-as-a-judge systems under rating indeterminacy","volume-title":"Target Int. J. Transl. Stud.","author":"Guerdan","year":"2026"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"197","DOI":"10.1075\/target.15.2.02hal","article-title":"The cognitive basis of translation universals","volume":"15","author":"Halverson","year":"2003","journal-title":"Target Int. J. Transl. Stud."},{"key":"ref14","article-title":"How good are GPT models at machine translation? A comprehensive evaluation","author":"Hendy","year":"2023"},{"key":"ref15","doi-asserted-by":"crossref","DOI":"10.4324\/9781315752839","volume-title":"Translation Quality Assessment: Past and Present","author":"House","year":"2014"},{"key":"ref16","first-page":"64","article-title":"Development of translation research in the context of large language models: issues and prospects","volume-title":"Chin. Transl. J.","author":"Hu","year":"2023"},{"key":"ref17","author":"Huang","year":"2025"},{"key":"ref18","volume-title":"arXiv preprint","author":"Jiang","year":"2023"},{"key":"ref19","article-title":"Is ChatGPT a good translator? Yes with GPT-4 as the engine","author":"Jiao","year":"2023"},{"key":"ref20","first-page":"419","author":"Karpinska","year":"2023"},{"key":"ref21","volume-title":"Rank Correlation Methods","author":"Kendall","year":"1962"},{"key":"ref22","first-page":"193","author":"Kocmi","year":"2023"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"159","DOI":"10.2307\/2529310","article-title":"The measurement of observer agreement for categorical data","volume":"33","author":"Landis","year":"1977","journal-title":"Biometrics"},{"key":"ref24","first-page":"13785","author":"Laskar","year":"2024"},{"key":"ref25","author":"Ma","year":"2026"},{"key":"ref26","doi-asserted-by":"publisher","first-page":"2035","DOI":"10.1007\/s11098-025-02347-3","article-title":"Normative conflicts and shallow AI alignment: R. Milli\u00e8re","volume":"182","author":"Milli\u00e8re","year":"2025","journal-title":"Philos. Stud."},{"key":"ref27","doi-asserted-by":"crossref","DOI":"10.4324\/9781315691862","volume-title":"Introducing Translation Studies: Theories and Applications","author":"Munday","year":"2016"},{"key":"ref28","first-page":"6423","author":"Naous","year":"2025"},{"key":"ref29","volume-title":"The Theory and Practice of [Biblical] Translation","author":"Nida","year":"1969"},{"key":"ref30","doi-asserted-by":"crossref","DOI":"10.4324\/9781315760506","volume-title":"Translating as a Purposeful Activity: Functionalist Approaches Explained","author":"Nord","year":"2014"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"27730","DOI":"10.52202\/068431-2011","article-title":"Training language models to follow instructions with human feedback","volume":"35","author":"Ouyang","year":"2022","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref32","first-page":"arXiv:2507.23009","article-title":"Stop evaluating AI with human tests, develop principled, AI-specific tests instead","author":"S\u00fchr","year":"2025"},{"key":"ref33","first-page":"283","author":"Torrielli","year":"2023"},{"key":"ref34","doi-asserted-by":"crossref","DOI":"10.1075\/btl.4","volume-title":"Descriptive Translation Studies\u2013and Beyond","author":"Toury","year":"1995"},{"key":"ref35","doi-asserted-by":"publisher","first-page":"427","DOI":"10.1080\/13556509.2024.2406103","article-title":"Political discourse translation in contemporary Chinese and western contents","volume":"30","author":"Valde\u00f3n","year":"2024","journal-title":"Translator"},{"key":"ref36","doi-asserted-by":"crossref","DOI":"10.4324\/9781315098746","volume-title":"The Translator\u2019s Invisibility: A History of Translation","author":"Venuti","year":"2017"},{"key":"ref37","first-page":"69","article-title":"Basic orientations for expanding translation studies","volume":"44","author":"Wang","year":"2021","journal-title":"Foreign Languages (Journal of Shanghai International Studies University)"},{"key":"ref38","first-page":"70","article-title":"Research on the innovation of translation education practice mode driven by large language model technology","volume":"45","author":"Wang","year":"2024","journal-title":"Chinese Transl. J."},{"key":"ref39","first-page":"27","article-title":"A study on the effectiveness of ChatGPT in translating discourse with Chinese characteristics","volume":"2","author":"Wen","year":"2024","journal-title":"Shanghai J. Transl."},{"key":"ref40","first-page":"72","article-title":"Reshaping translation studies: paradigm shift and future direction exploration under the influence of AI technology","volume":"47","author":"Yu","year":"2024","journal-title":"Foreign Languages (Journal of Shanghai International Studies University)"},{"key":"ref41","first-page":"68","article-title":"AI-empowered applied translation talent cultivation: exploring human-machine collaborative practice pathways","volume":"6","author":"Zhang","year":"2025","journal-title":"Shanghai J. Transl."},{"key":"ref42","first-page":"arXiv:2407.16891","article-title":"Cultural value differences of LLMs: prompt, language, and model size","author":"Zhong","year":"2024","journal-title":"arXiv preprint"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1869972\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T05:30:24Z","timestamp":1786339824000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2026.1869972\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,10]]},"references-count":42,"alternative-id":["10.3389\/frai.2026.1869972"],"URL":"https:\/\/doi.org\/10.3389\/frai.2026.1869972","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,10]]},"article-number":"1869972"}}