Perceptions of non-native speech are often guided by listeners’ expectations of a speaker. These expectations are informed by pre-existing beliefs about how particular types of people sound. Perceived ethnicity can affect how listeners evaluate speech (Rubin 1992; D’Onofrio 2019); however, most of this work has been situated in Western contexts. The current study details an experiment that tests for the linguistic profiling (Baugh 2005) of the Uyghur population of China, a group that has been systematically oppressed for their ethnicity and religion. Using name-based ethnicity priming, participants thought they were hearing either a Korean, Uyghur or non-descript speaker of L2 Mandarin. Results showed that participants rated the speaker as significantly more confident, intelligent and hard-working in the Uyghur condition. However, participants were significantly less likely to hire the supposedly ‘Uyghur’ speaker. We propose that these results are evidence of shifting standards (Biernat 2012), whereby listener expectations are lowered by social stereotypes, leading to inflated subjective ratings of minority groups, without leading to positive outcomes.