This study empirically investigates how interpreter ideology is manifested in the evaluative language of the World Economic Forum’s Annual Meeting in China in 2016 (English-Chinese language pair). Methodologically, van Dijk’s ‘Ideological Square’ and Martin & White’s Appraisal framework have been operationalised for the analysis of positive/negative evaluative language in ‘us’/’them’ discourses. The results reveal an overall positive-‘us’ and negative-‘them’ pattern in terms of interpreters’ ideological positioning. This manifested in three ways: i) negative, pejorative, and sensitive discourses about China have been self-censored, ii) positivity has been accentuated and negativity has been neutralised in China-related discourses, and iii) negative tones in the discourses of other countries have been amplified.