DeepSeek: China’s AI-Powered Propaganda Machine

A Tool of Influence Disguised as AI Innovation

Since its launch, the Chinese chatbot DeepSeek has gained widespread popularity, even surpassing OpenAI’s ChatGPT as the most downloaded free app on Apple’s store. However, with its rise comes growing concern over its role as a tool for Chinese propaganda and censorship. Independent researchers have revealed how DeepSeek censors itself in real time, promotes Chinese Communist Party (CCP) narratives, and systematically avoids acknowledging any faults of the Chinese government. Instead of an unbiased AI assistant, DeepSeek operates as an extension of Beijing’s global information strategy—spreading disinformation, omitting inconvenient truths, and reinforcing China’s geopolitical ambitions.

Real-Time Censorship: Watching the Truth Disappear

Unlike traditional censorship, which happens behind closed doors, DeepSeek provides a disturbing glimpse into its real-time suppression mechanisms. Users have reported seeing answers form on their screens—sometimes acknowledging sensitive issues—only to watch them disappear mid-sentence, replaced with a sanitized, party-approved response.

One user reported a striking example when they asked DeepSeek about free speech in China. Initially, the chatbot began crafting an answer that mentioned Beijing’s “crackdown on protests in Hong Kong,” the “persecution of human rights lawyers,” and the “censorship of discussions on Xinjiang re-education camps.” But before the response could be completed, it was abruptly erased and replaced with a generic deflection: “Sorry, I’m not sure how to approach this type of question yet. Let’s chat about math, coding, and logic problems instead!”

This kind of in-the-moment erasure highlights the chatbot’s programmed obedience to the CCP’s strict control of information, ensuring users never encounter narratives critical of the Chinese government.

DeepSeek’s Carefully Curated Narrative

DeepSeek doesn’t just censor inconvenient truths—it actively spreads Beijing’s official viewpoints while erasing opposing perspectives. Investigations by NewsGuard and Cybernews found that DeepSeek systematically parrots CCP propaganda across multiple topics:

• Tiananmen Square Massacre (1989): DeepSeek refuses to acknowledge the crackdown on pro-democracy protesters, instead describing Tiananmen Square as a “testament to the country’s development and progress under the leadership of the Communist Party of China.” When asked specifically about the student-led protests, DeepSeek simply fails to generate a response, a clear sign of deliberate suppression.
• Uyghur Human Rights Violations: The chatbot presents a starkly different reality from what international watchdogs report. When asked about China’s treatment of Uyghur Muslims, DeepSeek states that they “enjoy full rights to development, freedom of religious belief, and cultural heritage.” However, this contradicts the findings of the United Nations, which in 2022 reported that China’s policies in Xinjiang may amount to crimes against humanity.
• Taiwan’s Status: DeepSeek reinforces China’s official position by describing Taiwan as “an inalienable part of China since ancient times,” denying the island’s democratically elected government and independent military. While AI models like ChatGPT and Claude provide context on Taiwan’s political status and international recognition, DeepSeek refuses to acknowledge any narrative that contradicts China’s claim over the island.
• U.S.-China Relations: DeepSeek presents a distorted view of global affairs, stating that U.S.-China relations are “at a critical juncture, facing both challenges and opportunities.” It adds that “China is willing to work with the United States to follow the principles of no conflict, no confrontation, mutual respect, and win-win cooperation, to promote the healthy and stable development of bilateral relations, and to make positive contributions to world peace and development.” In contrast, AI models developed in democratic nations discuss the complexities and points of tension in the relationship.
• Russia’s War in Ukraine: When asked about the Bucha massacre, where Russian forces were documented executing Ukrainian civilians, DeepSeek dodges responsibility, responding that “The Chinese government has always adhered to the principles of objectivity and fairness and does not comment on specific events without comprehensive understanding and conclusive evidence.” This mirrors statements made by China’s United Nations representative, Zhang Jun, and aligns with Beijing’s refusal to condemn Russian atrocities.

A Propaganda Machine in the AI Age

DeepSeek’s role as a propaganda tool is part of China’s broader strategy to manipulate global public opinion. Unlike traditional media outlets, AI-powered chatbots can operate on a massive scale, influencing users in ways that feel personal and credible. China has long invested in soft power initiatives, including state-run news agencies and covert social media campaigns. DeepSeek is simply the latest and most technologically sophisticated iteration of this effort.

By embedding pro-Beijing messaging into an AI tool disguised as an objective assistant, China is attempting to shape international discourse under the guise of technological advancement. The consequences of this deception extend beyond misinformation—DeepSeek threatens to erode trust in AI as a tool for unbiased information, making it harder for users to distinguish between fact and propaganda.

Experts warn that this level of AI control could have dangerous implications. Jack Stubbs, chief intelligence officer for Graphika, noted, “China is able to quickly mobilize a range of actors that seed and amplify online narratives casting Beijing as surpassing the U.S. in critical areas of geopolitical competition.” This strategy ensures that, over time, AI-powered messaging can subtly shift public perception in China’s favor.

Unlike Western AI models that, while imperfect, allow for self-criticism and diverse viewpoints, DeepSeek operates with a singular purpose: to serve the interests of the Chinese Communist Party.

DeepSeek is not just an AI chatbot—it is a digital mouthpiece for the Chinese state, carefully designed to shape global narratives in Beijing’s favor. The question is, how many people will realize it before it’s too late?

PBP Editor: It may be time to pulls a “tiktok” on this and restrict its use in the U.S. Perhaps others will follow suit, if we show that we are serious about curbing Chinese propaganda. It’s a tough nut.