Written by Chu-hui (Emily) Yang.
Image credit: 鄭麗文/ Facebook.
In March 2025, The Wall Street Journal published a report discussing potential Chinese blockade strategies around Taiwan. A Taiwanese politician appeared to agree that China has the authority to encircle Taiwan. The video was shared on social media and gained public attention. However, I watched the full video on YouTube and realised it did not reflect the original context and meaning.
The video went viral for a while, but soon after, a Taiwanese news outlet published a fact-check article pointing out that the video had been edited and did not reflect the original meaning. I thought the discussion around this video would stop; however, surprisingly, the edited clips started appearing on social media again when the politician won the KMT chair election seven months later.
The video was shared on social media; even with a clarifying news article, it did not stop or slow down the circulation and discussion. The power of this manipulation did not come from advanced technology; simple video editing with misleading text, plus leveraging the credibility of a Taiwanese politician, could be powerful and effective in persuasion.
This case shows how information manipulation can work across different online environments, and how the same content can take on new political meanings as it travels.
It Starts from a Misleading Caption
The earliest indexable post is from a nationalist Weibo account. It shared the full clip with no edits, and the show title still appears, but with a misleading caption that frames former legislator Cheng Li-wun’s statement before viewers even watch the video, encouraging them to interpret her comments as endorsing the Chinese perspective rather than describing it as shown in the screenshot below.
Source: Weibo.
Amplification by State media
State-run media adopted the nationalist’s narrative and amplified it to a broader audience. Chinese media produced two versions of videos with very similar captions and accompanying text (see examples below). Both media-edited videos underscored the statement that Taiwan belongs to China, but with more edits. For instance, the media removed the original show title and presented the clips starting with Cheng saying Taiwan is part of China, making it seem like it was Cheng’s own view. The edited videos reinforced Beijing’s long-term messaging to justify its military measures around the island, emphasising that China’s actions comply with international law and are intended to protect Taiwan. After the videos were shared across multiple channels, they increased the visibility of these narratives.
Source: Watching the Taiwan Strait (Kan Taihai); Strait Herald.
Edited Narratives Cross Ideological Boundaries
Notably, the videos received high engagement not only from CCP-aligned accounts, but also from anti-CCP voices, who generally saw them as confirmation of their suspicions regarding the KMT’s position on sovereignty. On X, the most engaging post was from Teacher Li Is Not Your Teacher, a well-known and active anti-CCP account. He shared the video and commented, “Cheng stated: ‘Taiwan is part of China. China is encircling Taiwan today by deploying security units to protect Taiwan…’” This post gained over 569k views and 821 replies.
Only 3 comments mentioned that the video was edited, while most criticised Cheng as a traitor. The circulation of the misrepresented video through his account demonstrates that the pro-CCP narrative can also be influential in anti-CCP communities, without widespread awareness of manipulation.
Taiwanese media covered the manipulated video after it was widely shared on social media in March 2025, noting that it had been edited to make it appear as if Cheng had made the claim. Seven months later, the manipulated video once again drew online attention and criticism after Cheng won the chairperson election. Even after a clarifying report from a legitimate news outlet, the reframed video and statement kept circulating. Users continued sharing the clip, accusing Cheng of betraying Taiwan.
Some users also argued that Cheng had not publicly clarified her stance or criticised China for manipulating her remarks after it gained widespread attention. These posts portrayed her silence as implicit acceptance of the reframed narrative, further reinforcing criticism of both Cheng and, later, the KMT after she became party chair.
Critics later expanded their accusations from Cheng to the KMT, with users calling the party a CCP proxy. Another KMT legislator pointed out that the video was out of context and tried to smooth it over, but criticism continued. The manipulated video was no longer simply a piece of Chinese information manipulation; its political meaning evolved. Moreover, it continued to reinforce political polarisation.
The Power of Credibility Borrowing
The most interesting finding of this case is not the editing itself. It is the credibility the reframed version borrowed from the politician, which made the Beijing-aligned narrative appear more plausible.
The trajectory also shows why information manipulation does not necessarily depend on sophisticated technology. The edited video contains no AI-generated or deepfake features. A simple misleading caption was powerful enough to alter its meaning. State media then strengthened the narrative through selective editing and framing. Once the content entered other information environments, audiences interpreted it through their political assumptions.
This helps explain why correcting manipulated content can be challenging. A fact-check can establish that a video was edited, but the content may already have become useful for political arguments and shared on multiple platforms.
This case should not be taken as evidence that the actors or accounts involved were part of a coordinated operation. The earliest indexable post may not reflect the original source, and there is insufficient proof that the media outlets were conducting campaigns. The research also cannot determine whether engagement with the manipulated content had changed people’s political views.
However, the evidence shows a clear trajectory: a Taiwanese politician’s remarks were reframed through simple, misleading captions, then amplified by state-run media, circulated across platforms and political-ideology lines, and eventually re-entered Taiwan’s political discourse.
What This Case Tells Us About Information Manipulation
As Taiwanese political parties prepare for the upcoming local election, this case illustrates how Chinese information manipulation could interact with existing political divisions in Taiwan. Information operations are a form of grey-zone conflict that harms democracy without a physical attack. Once distorted information becomes available online, it feeds existing scepticism of Taiwanese politicians, political parties, and their supporters, and catalyses further partisan criticism.
However, politicians should take responsibility when others manipulate their statements. It’s crucial for the public to understand the position of elected officials and what they stand for. In Cheng’s case, her and the party’s lack of response leaves questions about her position unanswered and could deepen public distrust. Additionally, this case also sparked international conversations about Taiwan’s political sovereignty. Without a clear response, misleading interpretations may persist and affect how Taiwan’s international allies understand Taiwan.
The Taiwanese government should also take action to help the public access accurate information and prevent further social division. In the comments I reviewed, some users shared the full video, helping others recognise that the edited clip did not necessarily reflect Cheng’s own position. However, as mentioned before, most of the discussion continued to focus on attacking her and the KMT. Therefore, fact-checking articles are useful, but their reach could expand by sharing corrections across the platforms where the manipulated content circulates.
In the upcoming elections, it’s crucial for political parties to communicate their positions clearly to the voters and international audiences. More importantly, the public should have access to full, original statements and their context so that political messages are less likely to be misrepresented.
Information operations nowadays have become more AI-driven across different digital channels. Despite this, this case study demonstrates that a simple process can be as effective. The continued circulation of manipulated content suggests that narrative alignment may help misleading claims persist despite corrective information, allowing the state narrative to endure and contributing to political distrust.
Chu-hui (Emily) Yang works in a policy think tank in Washington, DC. She holds a Master of Mass Communication degree from the University of South Carolina and an M.A. in Journalism from National Chengchi University in Taipei. Her research focuses on Chinese information operations and political communication on social media.