The Fire and the Liar: Chinese disinformation about the Los Angeles wildfires
While wildfires have raged in Los Angeles since January 7, 2025, misinformation and disinformation have also been fueled in the worldwide media ecosystem. Like in many other global events, fake images and unfounded claims shared by English speakers were translated into Chinese and circulated on platforms favored by Chinese language users.
Among these false information pieces, AI-generated images of fire scenes were mostly popular. In addition, the false claims that inaccurately blamed DEI policies (Diversity, Equity, and Inclusion) for the California government’s failure to control the fires also found many Chinese-language speakers who took the opportunity to ridicule their alleged “leftist ideas.”
AI-generated videos went wild
After the wildfire broke out in Los Angeles, a significant amount of false information was propagated among Chinese language users on social media such as Facebook, Threads, X, and TikTok, including pictures and videos showing the rampant wildfires were devouring LA hills. Many of the images have also been shared widely in other languages on social media; some of them were recycled from old pictures, whereas several others were AI-generated. The images were not only used to depict the scenes of wildfires but were also opportunities for people who posted them to deliver religious or moral messages, expressing the belief that “God will protect those who have a religion or do good deeds.”
For example, a popular post featured pictures of an intact, red-roofed house standing alone among damaged blocks. The widely shared post noted that “the owner of the house was a religious person who was kind and often helped children in Africa” and used this case to demonstrate that God helped good and religious people. Nevertheless, these pictures were irrelevant to the current Los Angeles wildfires. The picture with green arrows pointing to the red-roofed house was actually an old photo shot at the scene of the 2023 Maui fires, and the birds-eye view image was generated by AI.
Other viral AI-generated pictures of the LA wildfires shared by Chinese-language social media users included the fake photo that the famous Hollywood sign was on fire and a video showing a palm tree-lined boulevard leading to a burning mountain.
Compared to prior natural disasters, we observed more AI-generated images and videos during the Los Angeles wildfires, indicating once again that AI technology has been one of the major challenges for the fact-checking profession. Another concerning trend is that AI is being used not just to display scenes that wrongly portray events but also to create propaganda videos, such as fake footage in which celebrities make untrue comments on current events while infused with ideas that the AI video creators wish to convey.
Fake celebrity AI video became a tool of propaganda
Among the widespread information during the LA wildfires, a few videos in which Elon Musk commented on the event piqued the interest of Chinese-language social media users, who circulated the videos and claimed that “Musk’s comments have marveled American people.” Unsurprisingly, the videos were AI-made, and Musk has never made the comments or produced the videos. These videos were originally shared by a YouTube page that was only created in December 2024. On the YouTube page, over a dozen videos show the AI Musk commenting on current events from the perspective of modern technology and warfare, including the possibility of a war between China and Taiwan. Given Musk’s reputation as an advanced technology entrepreneur who has been outspoken about current affairs, it seems reasonable to audiences that Musk makes such comments.
However, a closer look at “the Musk” in the videos reveals signs that the videos were made by AI. For example, Musk’s movements were repetitive, and most videos included excerpts from other videos to supplement the content. Taiwanese AI scientist Chih-Chung Hsu confirmed that the image of Musk in the video was created using deepfake technology. The Tawain FactCheck Center also found that the AI Musk part was copied from an episode of Lex Fridman‘s podcast program.
It’s unclear who built up the YouTube page and posted the videos. However, several of the YouTube videos were about China and Russia, and the narrative echoed themes often seen in Chinese propaganda, such as praise for the solidarity of Chinese people in times of crisis. In the aforementioned video in which Musk applauded China’s deployment of technology in the Tibet earthquake, the AI Musk also inserted messages complimenting China’s advantage in warfare, such as:
“If this were wartime, China’s response is what you’d call Blitz Krieg level efficiency, lightning war. They’d mobilize so fast it would make your head spin. Anyway, there’s a lot to admire here and a lot for us to learn…”
These videos have drawn hundreds of comments from audiences. A few audiences suspected that the videos were generated by AI. Still, others resonated with the messages in the video, applauding China’s people and government and criticizing the American government for being incapable and the American people for being selfish. It would be worth observing whether there will be more AI videos exploiting celebrities’ images to deliver narratives aligning with the messages promoted by China or other authoritarian countries.
Chinese language users echoed false information condemning DEI
In addition to AI images, another common narrative of false information about the LA wildfires in the Chinese language is the misrepresentation of California’s DEI policies. Those who created and spread the messages also inaccurately attributed the policies to the California government’s failure to control the fire in time.
Since the beginning of the LA wildfires, right-wing media and influencers, including Elon Musk, have criticized that resources have been overly allocated to DEI programs, which resulted in insufficient funding for fundamental needs. The criticisms have been disseminated on social media and intertwined with flawed assertions that need more context. Social media posts in the Chinese language also resonated with the criticisms from American influencers and media. For example, a Chinese social media post that attracted many engagements on Facebook claimed that:
“California firefighters need to use water stored in the reservoirs. However, left-wing environmental groups say that fish in the reservoirs should not be harmed. The California government thus crazily drained the water away and left the firefighters with no water to distinguish the fire.”
This claim was apparently contradictory (How could the government want to conserve fish while emptying water away from the reservoir?) but still received a lot of attention. Some elements of this claim could trace its origins to false statements in American right-wing politicians and media.
Another widespread false information was the hiring of Los Angeles Fire Department Chief Kristen Crowley and her work at the department. Chinese posts on social media claimed that “Crowley was hired because she is gay” and “Crowley aimed to hire more females or transgender persons because there were too many male firefighters.” These Chinese claims about Crowley and the LA Fire Department’s DEI policies echoed attacks from American influencers such as Collin Rugg, the founder of the conservative news website TrendingPoliticsNews.com. Rugg said in his X post that Crowley’s top priority in her work as the department chief was DEI.
This statement and view were also shared by Chinese users, who recounted them in the Chinese language and shared them on social media. Even though fact-checking organizations swiftly pointed out that Crowley never made this statement, the false information continued to spread widely among Chinese social media users and is viewed as an example of the “crazy leftists.”
This was not the first time that misleading information about DEI was extensively circulated among Chinese language users. In the 2024 Paris Olympic Games and the 2024 US presidential election, incorrect pieces about DEI issues originally posted in English were also shared in Chinese, spreading through platforms such as X and TikTok, influencing how Chinese-language users perceived DEI ideas and American politics. For example, the AI video of Kamala Harris stating she was a DEI hire was disseminated among Chinese-language social media users, influencing their opinions of Harris. The narratives promoted by American influencers and politicians have had an impact not only on English speakers but also on audiences in other language communities.
Wei-Ping Li is a research fellow at the Taiwan FactCheck Center