<![CDATA[In the digital age, AI has become essential in translation, providing fast solutions for overcoming language barriers. This is particularly important for the accurate translation of cultural elements within tourism promotional texts, which is crucial for achieving effective destination marketing. This study examines the translation result of ChatGPT and Google Translate, advanced AI-powered translation tools, in translating a tourism promotional text from the Jadesta Ministry of Tourism and Creative Economy, Indonesia. In detail, the study aims to answer any cultural aspects that are lost in the translation resulting from ChatGPT and Google Translate and to explore to what extent translation results favor source language orientation (foreignization) or target language orientation (domestication). Using cultural categorization from Chen (2024) to address translation loss and the concept of translation techniques from Molina and Albir (2002), a qualitative approach was applied to compare the translations of cultural references from ChatGPT and Google Translate. The steps of the study involved selecting a suitable tourism promotional text that contained culturally significant terms. The text was then translated using both ChatGPT and Google Translate, and the translations were evaluated based on their ability to convey cultural meaning. Expert validation was sought to ensure accuracy, followed by a qualitative analysis of the types and instances of cultural loss in each translation, leading to insights about the limitations of both tools in translating cultural terms. The findings reveal significant translation loss in terms of historical background, aesthetic imagery, local customs, and religion. Both ChatGPT and Google Translate show a cultural loss in translating local customs. Local custom terms are deeply ingrained in the source culture and often lack direct equivalents in the target language, making them particularly vulnerable to cultural loss during translation. Then, both tools predominantly employ pure borrowing techniques to preserve their cultural source and literal translation to ensure accuracy at the linguistic level but often overlook cultural and contextual values. In addition, both tools demonstrate a preference for source language orientation (foreignization). However, ChatGPT performs better than Google Translate due to its lower percentage of foreignization compared to Google Translate.]]>
