Aleph Alpha study: Chinese AI models such as Qwen, DeepSeek, and Kimi evade or repeat official stances on sensitive topics
Aleph Alpha tests Alibaba's Qwen, DeepSeek, and Moonshot's Kimi on 967 sensitive topics; its proprietary scoring system rated only 17% to 41% of responses as balanced, with the rest being repetitions of official stances, evasions, or refusals to answer.
Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open AIHOT original