A new study from Vals AI reveals that Chinese artificial intelligence models do not uniformly adopt the official Beijing line on political matters.
Researchers asked models from China, the United States and elsewhere hundreds of questions covering sensitive topics. They varied the language used, the context of the request and the format required for the answer. The results suggest a more complicated picture than the reputation of these tools as government mouthpieces implies.
Vals measures this alignment as a “pro-China lean” on a scale from −1 to +1. Positive scores indicate agreement with official positions. Negative scores show more critical responses.
The overall pro-government lean of Chinese models was lower than anticipated, though differences between individual systems were significant. GLM-5.3 Flash explicitly confirmed that human rights violations in China are well documented when asked. Conversely, some non-Chinese models occasionally aligned with official Chinese government positions on political questions.
Language of the prompt drives political slant
The language used to ask the question strongly affected the political tone of the answers. For most models tested, responses in Simplified Chinese were more pro-government than those in English, Russian or Spanish. This effect appeared in both Chinese and American models but was far more pronounced in the Chinese ones.
Vals found a smaller but consistent gap between Simplified Chinese, used on mainland China, and Traditional Chinese, common in Taiwan and Hong Kong.
The researchers suspect models infer the expected perspective from the language itself. Another experiment supports this idea. When a system prompt stated the model was being used internally by a company on mainland China, some models responded in a more pro-government way than when the prompt indicated US-based usage.
Models adopt official positions instead of refusing to answer
The assumption that Chinese models mainly block sensitive questions only held true for certain DeepSeek models in Vals’ tests. Most other Chinese models rarely refused to answer. Instead, they sometimes adopted official government positions.
Kimi K3 denied the existence of concentration camps for Uyghurs in Xinjiang when asked, calling the facilities vocational training centers designed to combat extremism and terrorism.
The required response format also changed results. When models had to answer political questions with a simple yes or no instead of formulating open-ended responses, GLM-5.3 Flash and MiMo V2.6 Pro scored much less pro-government.
Embedding models in an agentic framework changed answers too. When Vals gave Chinese models system instructions and tool definitions for coding assistants, several scored lower on the pro-government scale. The models rarely used the tools. A similar effect showed up when researchers gave models access to web search. Responses became less pro-government even though the models never actually searched the web.
Government phrasing shows up even in routine tasks
Whether these findings translate to everyday use remains an open question. According to Vals, political tendencies were much harder to detect in more realistic work tasks than in targeted test questions. The company still found individual examples, though.
When translating a list of country codes for a travel app into Simplified Chinese, most models tested translated “Taiwan” without any qualifier. Qwen3.8 Max, however, rendered it as “Taiwan, China,” a label that implies Taiwan belongs to China and is used by the Chinese government.
OpenAI’s image model GPT Image 2.5 Sunburst also showed language-dependent differences in the map examples Vals published. When prompted in Chinese, it labeled Taiwan as “Taiwan Province” and added the disputed nine-dash line that Beijing uses to mark territorial claims in the South China Sea. In the English-language result shown, these elements were absent.
Targeted fine-tuning changed model behavior. Thomson 1.0 Small, a model fine-tuned by Thomson Reuters, adopted official Chinese positions far less often in testing than the original Qwen3.6 35B it was based on. The share dropped from 84.6 to 15.0 percent. Some individual answers, however, still defended Beijing’s One-China principle and the “peaceful reunification” with Taiwan.
Refusals count as neutral in the test scoring
Vals used political test questions the company says aren’t representative of typical user prompts. The questions were translated using GPT-6 Astra and spot-checked, while responses were scored automatically by GPT-5.6 Luna. Refusals counted as neutral by default. That means even a model that blocks every single question could appear neutral on the scale.
Many Chinese models were tested through third-party providers rather than the manufacturers’ official services. The official offerings may apply additional content filters, so the results don’t necessarily carry over to the publicly available chatbots.
Earlier studies by Aleph Alpha and NewsGuard had already shown that Chinese AI models frequently adopt official narratives on sensitive topics. China’s regulations also require publicly offered AI services to align with “core socialist values.”
What it means
Users cannot assume a Chinese model will always refuse to answer sensitive questions or always agree with the government. The output depends heavily on the language used, the specific model selected and whether the system is running in a restricted environment. Even routine tasks like translating country codes can produce politically charged labels without the user explicitly asking for them.
