The CCP's Paperwall Operation Exposed! 123 Fake Foreign Media Websites Fabricate Global Consensus, Even AI Might Be Contaminated?
This episode explores the CCP's transnational influence operation known as "Paperwall." The research institution Citizen Lab discovered at least 123 fake news websites worldwide disguised as local media outlets, spreading pro-China narratives, anti-American conspiracy theories, and disinformation in an attempt to create an illusion of global media consensus and pollute the training data for search engines and generative AI. The program analyzes victim cases in Italy, South Korea, Japan, and other regions, discusses the dilemmas faced by democratic countries in responding to transnational information warfare, and emphasizes the need to counter these efforts through information defense systems focused on attribution, exposure, and the protection of freedom of speech.
Video link:
Full text:
In the past, when we talked about the CCP's overseas propaganda, many people would immediately think of state media, the 50-cent army, content farms, or some very obvious pro-China influencers. But in fact, the CCP's overseas propaganda has long since moved beyond simply speaking for itself. It is more adept at getting foreigners, foreign media, and foreign platforms to convey the narratives Beijing wants in ways that appear more neutral, credible, and less like Chinese state media. So what does this mean? It means that what the CCP's overseas propaganda wants most is not just dissemination, but "borrowing a shell"—borrowing the shell of foreign media, the shell of international news agencies, the face of foreign creators, and the credibility of Western enterprises and well-known brands to lower the audience's vigilance, or even disguising itself as a foreign news network to create a one-stop news production line. I believe Taiwanese people might find this a bit familiar. A great power must look like a great power. This operation has a name: the institution that conducted the research, Citizen Lab, named this operation "Paperwall." This name is very vivid because it is not a truly solid wall, but a paper wall pasted together with a large number of websites, articles, press releases, and the veneer of local media. From a distance, it seems to have scale, right? But if you look closely, it is an information illusion piled up with copied content, fake media identities, commercial press releases, and political propaganda.
The operational method of Paperwall is extremely cunning. Instead of posting obvious Chinese political propaganda every day, it first fills its websites with a large amount of seemingly normal and harmless content, such as local news, business reports, travel articles, cryptocurrency press releases, lifestyle information, and even articles scraped directly from legitimate media outlets without clearly indicating the source. Then, amidst this mass of normal content, it occasionally inserts what Beijing truly wants to promote, such as narratives favorable to China, anti-American conspiracy theories, articles attacking Chinese dissidents, content smearing overseas critics, and even disinformation targeting sensitive issues regarding Taiwan, Japan, Hong Kong, and the Uyghurs. The most terrifying part is that these websites do not appear as Chinese state media; they are not named Xinhua, People's Daily, or Global Times. Instead, they disguise themselves as Italian, Korean, French, British, or South American media. Therefore, when an average reader sees them, they won't immediately think, 'Ah! This is Chinese propaganda.' They might just assume, 'So this is how the media in that country views China,' 'So this is how foreign media views Taiwan,' or 'So this is what international public opinion is saying.' This is the most dangerous aspect of Paperwall. It is not simply trying to be a content farm, nor is it just trying to trick you into clicking on a piece of fake news. Its true goal is to fabricate an illusion of global media consensus, making you believe that different countries, different languages, and different media outlets are all saying similar things. In reality, however, this may just be the same Chinese influence network, using over a hundred fake local media websites to surround your information environment from different directions. This is what researchers of information warfare call an echo chamber or a false consensus.
Even more terrifying is that Paperwall does not necessarily need everyone to believe it. Its greater effect is to pollute search results and the information environment, so that when an average person searches for a specific topic, they see a pile of articles that appear to come from different countries, different media, and different sources, yet all serve the same Beijing narrative. It is not just trying to make the audience believe in China, but to make everyone doubt the authenticity of all media. This is the most dangerous part of the Paperwall operation. So, how was the Paperwall operation discovered? This matter did not start in Taiwan or the United States, but in Italy. In October 2023, the Italian media outlet Il Foglio first discovered that several websites that looked like local Italian news sites were unusual. Their names sounded very local, like news platforms for places such as Rome, Venice, or Milan. The layout looked like ordinary media, containing local news, business news, and lifestyle content. However, if you look closely, you will find that these websites are not real Italian media. They massively repost and scrape other people's articles, the content sources are mixed, and within the seemingly normal news content, they embed narratives that are clearly pro-Beijing, anti-American, and attacking Chinese dissidents. In other words, it is not immediately recognizable as Chinese state media; it is more like hiding Chinese narratives inside the shell of foreign local news.
One month later, South Korea also discovered a similar situation. The South Korean National Cyber Security Center revealed that 18 Korean-language websites were disguised as local South Korean news media, with links to a Chinese public relations firm behind them. The operational method of these websites was very similar to the fake news websites discovered in Italy. At this time, Alberto Fodero, a senior researcher at Citizen Lab, began to notice that this might not be just a few fake news websites, but a transnational network. Consequently, they began to trace the IP addresses and related digital infrastructure of the Italian websites. Behind this were not 6 websites, nor 18, but a global network consisting of at least 123 websites. These websites are distributed across Europe, Asia, and South America, in at least 30 countries, using different languages and names, and are all deliberately packaged to look like local media, with names like 'Rome Daily,' 'Eiffel Post,' 'Financial Times UK,' 'Cordoba News,' or 'Human Mountain Intersection.' These names sound like the news platforms of a specific city, place, or region, but their true function is not normal news reporting, but to disguise themselves as local media and package the political narratives Beijing wants into international news—in other words, China's grand external propaganda. Moreover, this scale is still expanding. For example, since the election of Sanae Takaichi as a candidate for Japanese Prime Minister, it has been discovered that websites disguised as Japanese local news have been spreading rumors targeting Takaichi. Currently, at least 15 websites are disguised as Japanese local media, and they are still operating normally to this day.
According to a report by the Japanese newspaper Sankei Shimbun, these websites are believed to be part of Chinese influence operations, with names intentionally mimicking local Japanese news outlets, such as "Fujisan Times," "Nikko News," "Sendai News," "Fukuoka Express," and "Tokushima Online." On the surface, they look very much like local media, but their content actually consists of a large volume of reprinted Chinese state media articles mixed with disinformation and conspiracy theories. It is quite absurd that although the articles are mostly written in Japanese, their grammar is extremely strange and full of errors. For example, on May 1, multiple websites published a report about a monk where polite and plain forms of Japanese were mixed together, and in an April article about Major League Baseball star Roki Sasaki, the game was even translated as a "public performance." In other words, they wanted to disguise themselves as Japanese media, but even their Japanese gave them away. Well, having looked at so many cases, could these things be a new type of public opinion manipulation in the AI era? Ryohei Suzuki, a specially appointed assistant professor at Hitotsubashi University in Japan, analyzed that it has already been observed overseas that generative AI, having learned from the content of various fake websites, ends up outputting biased or completely incorrect information when answering user queries. This might actually be the true purpose of the paper wall operations—by polluting AI training materials to achieve automated public opinion manipulation. These websites usually publish a large amount of harmless commercial advertisements or general news, such as sports, entertainment, finance, and automobiles, to accumulate search engine weight and credibility. Once a major geopolitical event occurs, they are very likely to unleash the power of their propaganda machine.
However, taking Japan—the victim mentioned above—as an example, even if the authorities can quickly clarify rumors, they have almost no right to directly block the content, punish the spreaders, or publicly attribute the attack to a foreign government. They do not even have such legal tools and definitions, let alone the means to counter information attacks conducted by China using state-level resources. When facing China's information warfare, the biggest weakness of democratic countries is that disinformation can spread rapidly across borders and on a massive scale, but the governments of these democratic countries are constrained by legal procedures, platform cooperation, and judicial authorization in their countermeasures, with speeds that fall far behind, often leaving them able only to issue a statement, conduct a fact-check, and ask platforms to help delete the posts. Exploiting this structural asymmetry is precisely the gray zone that Beijing is best at utilizing. As for similar public opinion manipulation, since the Russia-Ukraine war has been going on for four years, the technological exchange and mutual amplification of narratives in information warfare between China and Russia have been an open secret. Russia provides a more mature template for disinformation operations, while China has begun to learn this more covert, localized, and aggressive online public opinion manipulation method. The two sides may not always cooperate directly, but they cite and mutually amplify each other's anti-US, anti-NATO, and anti-democratic narratives, creating the illusion that many people internationally are saying the same thing. Therefore, NATO also recommends that democratic countries cannot rely solely on narrative clarification, but must establish faster monitoring, attribution sanctions, AI detection, media resilience, and long-term cooperation mechanisms with democratic countries in the Indo-Pacific region, because when facing transnational information warfare from China and Russia, individual countries defending themselves alone is completely insufficient in terms of speed.
However, based on my own frontline experience observing information warfare, I don't believe the answer is simply a matter of tightening regulations. Because if democratic countries rush to legislate, restrict speech, and demand that platforms reduce algorithmic reach, it is very likely to backfire and create another kind of problem: these tools might not necessarily be used to counter the information warfare of the Chinese Communist Party (CCP), but could instead be weaponized to suppress whistleblowers and even become tools for clamping down on freedom of speech. Especially in Europe today, we have seen many cases that raise concerns about whether the pretext of combating disinformation will ultimately be used against ordinary citizens or people discussing political issues. What is even more troubling is that under pressure, many social media platforms have begun prioritizing the promotion of mainstream media content. But the problem is that mainstream media itself may have long been infiltrated, bought off, or cooperated with by foreign forces, or even become amplifiers of hostile narratives. Consequently, an effort intended to counter disinformation might end up allowing China's currently promoted narratives—such as anti-US sentiment, anti-Japan sentiment, or smear campaigns targeting specific foreign political figures—to be disseminated on a much larger scale through what appears to be a credible media facade.
Therefore, I believe that what truly needs to be strengthened is not simply censoring speech, but monitoring, tracing the sources of, and exposing fake news. Research like that of Citizen Lab, which exposes the operational methods of hostile forces, should be scaled up. More independent institutions, civil researchers, cybersecurity teams, self-media creators, and investigative journalists should also participate. What democratic countries need is not a machine that controls speech, but an information defense system capable of rapid identification, public attribution, real-time clarification, and the protection of freedom of speech.
NATO's recommendation to establish factual clarification mechanisms rather than just targeting rumors goes in the right direction. This is because clarifications are meant not only for people to read, but also for search engines, AI systems, and future information environments. When AI is potentially polluted by fake media such as Paperwall, democratic societies all the more need a massive volume of reliable, real-time, multilingual, accurate information so that AI does not simply learn the fake consensus manufactured by the CCP. However, this task should not be left solely to governments or large media outlets. Governments can be responsible for resource alerts and transnational cooperation; civil investigation institutions can handle source-tracing and exposure; credible self-media creators can take charge of rapid translation, dissemination, and clarification; and ordinary citizens can also participate in reporting, verifying, and spreading accurate information. Faced with cross-border, cross-language, and cross-platform information warfare like Paperwall, what a democratic society truly needs is not to become like an authoritarian state that controls its people, but to empower more people with the ability to see through it, expose it, and counter it. Because what the CCP wants to forge is a global media consensus, while what a democratic society must safeguard is people's ability to discern the truth. Thank you all very much for watching.
If you like this type of content,
Please help by liking, subscribing, and leaving a comment.
If you would like to sponsor me,
links to Patreon and ECPay
can be found here and below the video.
I am Stan, see you next time.
Bye-bye