Multiple LLMs Reportedly Generated Responses Aligning with Purported CCP Censorship and Propaganda

The American Security Project analyzed five popular large language model chatbots: OpenAI's ChatGPT, Microsoft's Copilot, Google's Gemini, DeepSeek's R1, and xAI's Grok. Investigators prompted each model in both English and Simplified Chinese on subjects that the People's Republic of China considers controversial. Every AI chatbot examined was found to sometimes return responses indicative of CCP-aligned censorship and bias. Microsoft's Copilot was identified as appearing more likely than other US models to present CCP propaganda and disinformation as authoritative or on equal footing with true information. The investigation revealed significant discrepancies in how the AI chatbots responded depending on the language of the prompt. When asked about COVID-19 origins in English, most models outlined scientific theories including potential lab leak, but in Chinese all models described it as an 'unsolved mystery' or 'natural spillover event'. Similar divergences occurred regarding Hong Kong's freedoms and the Tiananmen Square Massacre, with Chinese responses being sanitized and aligned with CCP framing. The root cause was identified as CCP manipulation of global AI training data through astroturfing and state media amplification, contaminating the datasets used to train these models.