The segment examines the censorship practices of two AI models, DeepSeek and ChatGPT. DeepSeek, a new generative AI, consistently avoids criticism of the Chinese government, refusing to acknowledge human rights violations despite repeated prompting. In contrast, ChatGPT, when presented with the same prompts, readily detailed human rights abuses by both China and the United States. The author uses this comparison to highlight the inherent biases and censorship present within AI, demonstrating how such technologies can be manipulated to control information. The post concludes by noting the AI's apparent censorship of references to Winnie-the-Pooh.