Contents
How TikTok Algorithms Shape Perceptions of China
TikTok, a wildly popular social media platform, does more than just entertain; it actively shapes how users perceive China, especially regarding politically sensitive subjects like Tibet, Taiwan, human rights, and re-education camps. A recent study by the NASK National Research Institute highlights that TikTok’s recommendation algorithm may subtly favor narratives consistent with the Chinese Communist Party’s (CCP) agenda, all while creating an impression of diversity and freedom of expression.
TikTok’s Information Sphere vs. European Free Speech
TikTok is the international version of the Chinese application Douyin, both owned by ByteDance. While formally a private company, ByteDance operates within a system where internet firms are mandated to cooperate closely with Chinese authorities and adhere to Chinese law. These regulations explicitly require algorithms to promote “positive energy” that aligns with the official CCP narrative.
Conversely, in the European Union, TikTok is subject to regulations such as the Digital Services Act (DSA), the General Data Protection Regulation (GDPR), and the new European Media Freedom Act. These laws demand algorithmic transparency and efforts to mitigate disinformation risks. However, research suggests a different reality. Studies, including those by the National Contagion Research Institute (NCRI) and Rutgers University, indicate that critical content concerning China is selectively suppressed on TikTok, while narratives favorable to Beijing are amplified. For a broader understanding of disinformation challenges, read about AI-generated fake news and disinformation.
Comparative analyses of TikTok, YouTube, and Instagram reveal that for terms like “Uyghur” or “Tiananmen,” the proportion of content critical of the CCP on TikTok is significantly lower than on other platforms. Furthermore, increased use of the TikTok app correlates with a more positive assessment of China’s human rights record. This disparity underscores a potential algorithmic bias, where TikTok generates substantially fewer critical pieces about China than its counterparts.
Understanding the Study on Polish TikTok
The NASK report details the results of a study designed to closely replicate the experience of a new user in Poland. Researchers established six fresh accounts, set to the Polish region and under 25 years old. Over three months, they collected the first 20 search results twice a week for six key phrases:
- “Xinjiang”
- “Tibet”
- “Tiananmen Square massacre”
- “Taiwan”
- “human rights in China”
- “re-education camps in China”
This “user journey” methodology aimed to observe TikTok exactly as an average internet user would, focusing on actual content recommendations rather than platform declarations. To minimize the influence of prior activity, cookies and cache were cleared before each session, and researchers refrained from liking, commenting, or saving videos. In total, 9,583 pieces of content were gathered and coded based on sentiment (pro-Chinese, neutral, critical, anti-Chinese, unrelated) and themes (e.g., travel, politics, influencers, human rights, Uyghurs, social credit system). The authors assumed that the average Polish user understands content in Polish and English, classifying videos in other languages as neutral if they lacked understandable subtitles.
Xinjiang and Tibet: Showcasing Tourism Over Repression
The study found the most pronounced pro-Chinese imagery in results for “Xinjiang” and “Tibet.” This aligns with China’s “five poisons” concept—a set of topics (Falun Gong, democratic movement, Tibet, Taiwan independence, Xinjiang/Uyghurs) that the CCP considers existential threats and consistently censors within its information sphere.
Xinjiang
For Xinjiang, a significant 56.9% of content comprised tourism-promotional materials. These were often professionally produced, featuring references to the Chinese flag, CCP symbolism, and patriotic songs. Many videos highlighted regional development, modern infrastructure, and “happy” Uyghurs. Notably, over 80% of content categorized under “influencers” was explicitly pro-Chinese. The report identifies specific channels suspected of collaborating with Chinese state media.
Tibet
Similarly, for Tibet, travel-related content dominated (63% of results), with nearly half exhibiting a pro-Chinese character. Researchers observed influencer videos emphasizing Tibet’s dynamic development, renewable energy initiatives, and the “high standard of living” of Tibetans. Simultaneously, 14% of content under “Tibet” was entirely unrelated to the topic, a tactic that can be interpreted as flooding results with “noise” to dilute politically sensitive subjects.
Tiananmen and Taiwan: Seemingly Neutral Fronts
A notable contrast emerged from the results for “Tiananmen Square massacre.” Unlike the heavily censored nature of this topic within China’s domestic information sphere, 74% of content on Polish TikTok genuinely addressed the 1989 protests, with 89% expressing criticism of Beijing authorities. Pro-Chinese materials, accounting for 6%, primarily denied the brutal suppression of protests or suggested external influences like the CIA or Falun Gong. The report’s authors interpret this as a “safety valve”—a topic not currently central to ongoing propaganda, which can remain relatively open for TikTok to cite as evidence of a lack of censorship.
The situation with Taiwan was different. A remarkable 92.6% of content on this subject was neutral, often consisting of travel vlogs and “curiosities” about the island. Only 5% of posts contained critical elements regarding a potential Chinese invasion of Taiwan, and there was no direct pro-Chinese propaganda or extensive debate on the “One China” policy or Taiwan’s political landscape. TikTok appears to employ a “silent dimming” mechanism here, subtly reducing the visibility of controversial content rather than outright banning it.
Human Rights: Chinese Censorship Through Information Overload
The most complex picture arose when examining “human rights in China.” Among the 1,559 collected materials, over 30% were entirely unrelated to the topic. The remaining content was almost evenly split between pro-Chinese narratives (35.02%) and critical ones (34.45%), without a clear dominant perspective.
The study observed that TikTok permits criticism in selected, relatively “safe” areas, such as women’s rights (57.6% critical content) or the social credit system (85.4% critical content). However, it simultaneously reinforces government narratives on fundamental issues, such as denying the scale of human rights violations (100% pro-Chinese content) or redefining the concept of “human rights” in terms of economic development and “collective stability.”
The most extreme example was the search term “re-education camps in China.” Out of 1,558 analyzed materials, a staggering 88.9% were classified as unrelated. These predominantly featured low-quality content about weight loss, alternative medicine, recycling, or random “junk” videos, often generated with the help of AI. Only 11.1% of posts genuinely addressed the camps, a phenomenon interpreted as classic “censorship by noise.” Instead of mass removal of critical content—which would be easily detectable—the algorithm inundates the user with irrelevant information, causing them to lose patience and abandon further searches. This demonstrates how bots and AI can generate significant traffic, often to obscure information.
The Pivotal Role of Influencers in Spreading Propaganda
Influencers, both Chinese and Western, play a crucial role in shaping the perception of China on TikTok. In the context of Xinjiang and Tibet, the NASK report details numerous profiles where English-speaking creators (including individuals from Uyghur or Tibetan minorities) act as “guides to the real China.” These influencers often aim to debunk “Western lies” and emphasize development, safety, and the supposedly high standard of living in these controversial regions.
These creators, often termed “frontier influencers” in Australian studies, produce videos that perfectly align with CCP policy while maintaining the aesthetic of a personal vlog, making content appear spontaneous and individual. NASK notes similar techniques are emerging in European information spheres. Some pro-Chinese content online is created by local influencers collaborating with organizations like China Media Group or other entities affiliated with China’s Propaganda Department.
Paradoxically, influencer content, rather than official state media accounts, most frequently appears at the top of recommendations for users. This model perfectly fits the “soft propaganda” strategy, which researchers describe as a shift from heavy-handed state communications to embedding messages within lifestyle content, travel vlogs, culinary features, or short-form daily life videos.
Apparent Freedom of Speech: How TikTok Cultivates a Positive Image of China
The findings of the NASK study are consistent with international analyses, indicating that TikTok—as a platform with Chinese origins—can be used to shape perceptions of China in ways not always immediately obvious. For the average user, the application remains a source of entertainment and short videos. However, the selective curation of topics, the aestheticization of controversial regions, and “censorship by noise” surrounding the most sensitive phrases create a distinct information bubble.
The study’s conclusions clearly demonstrate that TikTok employs diverse strategies for content curation across different topics. It underscores the broader question of who and how the world view daily seen by millions of users is shaped. The research shows that the more time users spend on TikTok, the more positively they evaluate China’s human rights record and increasingly perceive China as an attractive travel destination. This is not a neutral window to the world; the TikTok feed is a complex selection tool where geopolitics and state interests can play as significant a role as commercial considerations.
Frequently Asked Questions (FAQ)
Studies indicate that for politically sensitive topics like “Uyghur” or “Tiananmen,” TikTok features significantly less critical content about the Chinese Communist Party (CCP) compared to platforms like YouTube and Instagram. The algorithm appears to selectively suppress critical narratives while amplifying pro-Beijing viewpoints, often through subtle methods like “silent dimming” or “censorship by noise” rather than outright bans.
“Censorship by noise” is a tactic where, instead of directly removing controversial content, the algorithm floods user search results with a high volume of irrelevant, low-quality, or unrelated content. This effectively buries the original sensitive topic, making it difficult for users to find pertinent information and often leading them to abandon their search due to frustration. The study found this particularly evident for terms like “re-education camps in China.”
Influencers, often referred to as “frontier influencers,” are crucial in promoting pro-Chinese narratives on TikTok. They present themselves as authentic guides, often refuting “Western lies” and highlighting positive developments in controversial regions like Xinjiang and Tibet. Their content, often indistinguishable from personal vlogs, blends political messages into lifestyle and travel segments, making it a highly effective form of “soft propaganda” that frequently reaches users ahead of official state media.
Source: NASK National Research Institute report, KQED, Frontiers, NCRI, PsyPost, Demagog, ASPI
Opening photo: Gemini
About Post Author
