China’s DeepSeek AI Faces Outrage over Uyghur Censorship and Propaganda

BEIJING: Human rights activists and international experts have expressed strong criticism of China’s newly developed AI platform, DeepSeek. Concerns have been raised about its role in disseminating state propaganda, censoring sensitive topics, and collecting personal data. Although marketed as a technological breakthrough, the platform has encountered significant ethical and security concerns, according to the Campaign for Uyghur.

DeepSeek is reported to censor discussions about the Uyghurs, the Tiananmen Square Massacre, and calls for Taiwan’s independence. In addition, it collects a wide range of personal information including IP addresses and chat histories, which are stored on servers located in China.

The Campaign for Uyghur has highlighted that DeepSeek aggressively gathers and retains sensitive personal data, raising fears of potential misuse by the Chinese Communist Party (CCP), a regime notorious for its history of human rights abuses. The platform has also been accused of silencing dissent, particularly regarding Xinjiang, leading to concerns that it acts as a tool for state-sponsored digital surveillance.

Rushan Abbas, the Executive Director of the Campaign for Uyghurs, condemned the platform on social media, stating, “It collects sensitive data that would benefit the CCP, a regime known for human rights abuses. Chinese AI platforms and apps represent threats including digital transnational repression. We cannot overlook this issue.” Her remarks underscore the serious concerns about AI tools that can facilitate censorship and surveillance.

Jan Czarnocki, a digital law expert from Switzerland, shared his experience with DeepSeek on X, stating, “DeepSeek might be a technological advancement, but it serves as a mouthpiece for Chinese propaganda.” He attempted to discuss the actions of China in Xinjiang, hoping to have the platform acknowledge them as genocide and crimes against humanity. However, each time he brought up the issue, the AI ceased to respond, claiming that the query could not be addressed. Czarnocki described the AI’s responses as heavily predetermined, labeling them “a textbook example of Chinese propaganda.”

Czarnocki proposed that DeepSeek's open-source framework should be retrained using truthful information to enhance security and reduce reliance on the Chinese API.

Further criticisms emerged from Open Source Intel, a US-based news and intelligence organization, revealing that when they posed questions about the treatment of Uyghurs to DeepSeek, it failed to provide substantive answers.

Critics assert that platforms like DeepSeek pose a growing threat to free speech and privacy. They are urging the international community to take a strong stance against such tools of repression.

China, AI, Uyghur