By Wei-Ping Li, Ph.D.

Since last December, a wave of AI-generated videos praising Taiwan has emerged across various YouTube channels. Initially, these videos appear clichéd, showcasing compliments on the kindness and integrity of the Taiwanese people, highlighting Taiwan’s medical progress, and criticizing the moral decline and economic downturn in China. Recently, some channels have begun sharing stories seemingly opposing a potential Chinese invasion and supporting Taiwan’s participation in international organizations. These videos utilize AI-generated personalities to promote Taiwan and fabricate news events that have never occurred. While it is difficult to discern the intent behind these AI-generated videos, many of the videos have gone viral and have even been featured in Chinese-language mainstream media news coverage. 

The AI-video series of “Chinese girls amazed by the great things in Taiwan’s society”

The AI-generated video that first attracted attention was titled “Why are Taiwanese not Chinese? A girl from Shenzhen was shocked after traveling to Taiwan. The gap [between Taiwan and China] is too big. [為什麼台灣人不是中國人,深圳女孩旅遊台灣後震驚,真實差距太大]” In this video, a girl named “Xiaoqian [小茜],” who claimed to be a newly graduated master’s student from Shenzhen, China, shared her thoughts on her recent visit to Taiwan, comparing the salaries and benefits received by young individuals in Taiwan and China. She was saddened by the low incomes and bureaucracy in China while praising Taiwan’s freedom, as well as the Taiwanese people’s care for one another. In the end, the girl, who self-identified as a former “little pink” (a Chinese nationalist), concluded that she eventually realized the difference between China and Taiwan. A YouTube account named “Database for Videos Taken Down in Seconds [中國秒刪影片庫]” was one of the channels that shared this video in late December 2024. Soon, this video was picked up by multiple YouTube channels and other social media platforms, as well as Taiwanese TV news such as SETN and overseas Chinese media, New Tang Dynasty TV.

This video is not authentic, though. Experts interviewed by the Taiwan FactCheck Center confirmed that it was generated by AI, and the girl “XiaoQian” was actually a lifelike AI avatar. Interestingly, multiple similar videos on this topic surfaced on various YouTube channels around the same time. Some AI-made girls in the videos claimed to be from Beijing, while others stated they were living in Shanghai, all expressing their astonishment at the politeness of the Taiwanese people and the effectiveness of the Taiwanese government. 

A screenshot of online news coverage on the "Shenzhen girl's Taiwan story" by a Taiwanese media outlet.
A screenshot of online news coverage on the “Shenzhen girl’s Taiwan story” by a Taiwanese media outlet.

Following the series “Chinese girls amazed by the great things in Taiwan’s society,” YouTube channels promoting Taiwan’s excellence have sprouted over the past few months. The avatars featured in the videos vary, and the scripts differ, but they share similar formats and themes. Another popular theme, aside from the series about Chinese individuals being amazed by Taiwan, is “foreigners admire Taiwan’s life,” which highlights Taiwan’s advancements in medical treatments. The foreigners who appreciated these medical advancements were often nobles from Europe or the Middle East who had suffered from severe diseases that could not be treated in their own countries. They sought Taiwanese doctors who integrate Western and traditional Chinese treatments to save lives. Unsurprisingly, these foreign nobles were miraculously cured by Taiwanese doctors. Like the series featuring Chinese individuals, the “foreign noble series” is also narrated by AI avatars against a backdrop of repeated scenic or street views irrelevant to the stories.

Fabricated diplomatic events

Since April, these video channels have also expanded their subjects to political issues and highlighted how democratic countries have advocated for Taiwan and criticized Chinese diplomats at international diplomatic events. One of the earliest videos debuted in April, when the tariff war between Trump and China began. This fake video alleged that an official from the European Union, whose name was translated into Chinese as “雅各·范德倫” (pronounced similarly to “Jacob van der Linden”), delivered a speech at a United Nations conference, bringing attention to the rise of Taiwan amid the trade conflict between the United States and China. The EU official even boldly predicted that Taiwan would join the United Nations and potentially become a permanent member of the United Nations Security Council. In May, more videos depicting fabricated diplomatic events were uploaded to the YouTube channels. One of the most frequently shared stories featured a UN meeting in which a Japanese representative defended Taiwan and chastised an incensed Chinese wolf warrior diplomat who insisted that Taiwan is a part of China. Videos in this and other channels with similar narratives substituted representatives from Japan with diplomats from other countries, including Germany and Australia. 

Screenshots of YouTube videos claiming that representatives from countries such as Japan, Germany, and Australia supported Taiwan's entry into international organizations or defended Taiwan's position by reprimanding Chinese diplomats.
Screenshots of YouTube videos claiming that representatives from countries such as Japan, Germany, and Australia supported Taiwan’s entry into international organizations or defended Taiwan’s position by reprimanding Chinese diplomats.

In addition to using AI-generated avatars to tell the stories, the events and names of the diplomats in these videos were also fabricated. For example, the EU diplomat “Jacob van der Linden” does not exist. Furthermore, the image of Jacob van der Linden and the scene of him giving a speech were entirely produced by AI. Another scene in this video, while verified to be a real UN event, was actually an occasion where the Secretary-General of the United Nations, António Guterres, delivered a speech to commemorate the genocide against the Tutsi in Rwanda. Similarly, the videos featuring diplomats in Japan and other countries were created by AI tools to portray fictional events. The diplomats’ names in these videos, including those of the UN ambassadors from Japan and China, are not real and even vary across different YouTube channels that present the same story.  

A screenshot of an AI-generated video that claimed "an EU senior consultant spoke in the UN, saying Taiwan will become one of the top three countries and will become a permanent member of the UN Security Council."
A screenshot of an AI-generated video that claimed “an EU senior consultant spoke in the UN, saying Taiwan will become one of the top three countries and will become a permanent member of the UN Security Council.”

Characteristics of these pro-Taiwan AI videos

As of now, over a dozen YouTube channels featuring AI-generated pro-Taiwan videos have been identified, though there may be more. These channels and their videos exhibit various common traits.

The self-claimed identity of the YouTubers/Narrators

The majority of the AI avatars narrating the story in the video are young women. Many provide vague background stories in their channels’ introductions. Some of them identify themselves as tour guides, travelers, or former flight attendants from Korea, Japan, or China, who fell in love with Taiwan during their visits to the island. A few of them used similar phrases and emoji combinations in their introductions to encourage audiences to email their contacts, stating, “😊 Visit Taiwan, learn about Taiwan, fall in love with Taiwan!” and “Don’t forget to subscribe, comment, and share! 💬😊.” Among the accounts we analyzed, the earliest one was created in 2020, but all began posting content after December 2024. These channels have accumulated thousands of subscribers, with certain videos gaining even more traction. One example is the video about the Shenzhen girl’s reflections after her trip to Taiwan, which has garnered over 440,000 views since December 2024. 

Regarding the avatars’ features, while many covered their faces with masks, some displayed their faces. Some avatars even claimed affiliation with renowned institutions. In some cases, one avatar could appear in multiple videos with various identities. For example, in one of these videos, an avatar introduced himself as Alan Thompson, an English citizen who hosted a documentary program for the BBC. In reality, there was indeed a BBC host named Alan Thompson. However, the actual Alan Thompson passed away in 2017 and bore no resemblance to the AI avatar in the fake video. 

Screenshots of AI-generated avatars in videos promoting Taiwan with false claims.
Screenshots of AI-generated avatars in videos promoting Taiwan with false claims.
Screenshots of AI videos from the same YouTube channel. The same avatar was repeatedly used in different videos, each with a different identity and name. The one on the right even claimed to be Alan Thompson, an English citizen working for the BBC as a host for a documentary program. The real BBC host, Alan Thompson, who had already passed away in 2017, looked completely different from the AI avatar in the fake videos.
Screenshots of AI videos from the same YouTube channel. The same avatar was repeatedly used in different videos, each with a different identity and name. The one on the right even claimed to be Alan Thompson, an English citizen working for the BBC as a host for a documentary program. The real BBC host, Alan Thompson, who had already passed away in 2017, looked completely different from the AI avatar in the fake videos.

The narratives in the AI-generated videos

The overarching theme of these videos is to praise Taiwan and advocate for it, but many narratives are based on comparisons between Taiwan and China, particularly emphasizing the moral qualities of Taiwanese people (as friendly and honest) and Chinese people (as morally corrupt or greedy). The revelation in the stories often comes from a Chinese or foreign visitor’s traveling experience to Taiwan, who eventually realizes that Taiwan is an emblem of lost virtues in current Chinese society. As a matter of fact, this storyline is not new and has been frequently reiterated in Taiwanese media to appeal to Taiwanese audiences since the 2010s, when the famous Chinese writer Han Han published an article, “The Wind from the Pacific Ocean [太平洋的風],”  after his visit to Taiwan, which praised Taiwanese kindness while lamenting the erosion of integrity in a materialistic Chinese society. 

Additionally, as previously mentioned, these YouTube channels also provide timely updates on current international events or incidents in Taiwan, such as the controversy surrounding Taiwan’s deportation of Chinese spouses who advocated for using military force to unify Taiwan, Nvidia’s decision to establish a new overseas headquarters in Taiwan, and the recent annual pilgrimage honoring the Taoist sea goddess, Mazu. Some of the videos also aim to boost Taiwanese confidence, suggesting that China still lacks the capability to invade Taiwan and that the international community largely supports Taiwan. However, as this analysis shows, many events depicted in the videos never actually occurred. 

The intent and the dissemination of the videos

Determining the intent behind false information is often challenging. While the videos commend Taiwanese people and society, promote Taiwan’s international standing, and criticize China’s ambition to invade Taiwan, they seem to target overseas Chinese audiences curious about Taiwan, as well as Taiwanese individuals who appreciate these “feel-good” narratives and are willing to contribute to streaming revenues or even donate. Notably, a few YouTubers have cautioned against such AI-generated videos, warning that they might be merely cheap fakes designed to profit from audience trust by promoting fabricated information. 

Regardless of what has motivated content creators to produce AI-generated videos, these AI videos have captured the attention of audiences who have further shared some of the videos on X and Facebook. A few fake videos were even amplified by mainstream Taiwanese media as genuine events, attracting more clicks on these videos. Taiwanese audiences and journalists should remain vigilant about the use of online news sources and exercise greater caution when evaluating the credibility of information.   

Wei-Ping Li is a research fellow at the Taiwan FactCheck Center. 

Rogge Chen (fact-checker at the Taiwan FactCheck Center) contributed to this analysis.