By Macrina Wang and Elisa Xu | Published on Sept. 11, 2023

Additional reporting by Sara Badilini and Natalie Huet

Dozens of pro-China social media and blog accounts are spreading the false claim that an experimental U.S. military weapon caused the August 2023 Maui wildfires — in what appears to be part of a coordinated online campaign to portray the U.S. military in a negative light, NewsGuard has found.

In August and September 2023, NewsGuard identified 85 social media and blog accounts spreading nearly identical posts and videos claiming that British intelligence service MI6 had revealed that the U.S. caused the Hawaiian wildfires, which left at least 115 people dead. This content was found on 14 major platforms, including Facebook, X (formerly Twitter), YouTube, and blog posting site Medium.

“This Hawaiian wildfire is just a ‘weather weapon’ attack experiment conducted by the US military!” declared many of the accounts in the campaign.

The influence operation appears to have been launched by Chinese speakers. NewsGuard was unable to determine whether the Chinese government was directly involved. However, this is not the first pro-China influence operation NewsGuard has identified. In April, NewsGuard detected a disinformation campaign on X disparaging two Chinese dissidents, and in November 2022, NewsGuard discovered a campaign, also on X, targeting an NGO that published a critical investigation of Beijing’s overseas influence.

This newly-discovered network appeared to be aimed at users in multiple countries, posting in 15 languages in addition to Chinese — English, Korean, Russian, French, Italian, German, Japanese, Indonesian, Dutch, Icelandic, Filipino, Maltese, Belarusian, Malagasy, and Marathi, one of India’s official languages.

Posing as ordinary social media users, some of these accounts reposted and interacted with one another, artificially inflating the network’s reach. Many accounts also used hashtags such as #meteorologicalweapon to attract attention. (Some of the platforms used by this campaign, such as Facebook and X, have policies against coordinated inauthentic behavior.)

Asked about NewsGuard’s findings, Meta said in a September 2023 email that some Facebook accounts shared by NewsGuard are part of a spam operation originating from China that Meta has been monitoring since 2019. Similarly, a Reddit spokesperson said that the Reddit accounts showed patterns of behavior associated with the same spam operation.

Medium told NewsGuard that the Medium accounts violated the platform’s policy against misinformation and inauthentic activity.

NewsGuard also contacted the other major platforms targeted by the influence operation — X, YouTube, Quora, Vimeo, Tumblr, Pinterest, Tripadvisor, Blogger, Rumble, story platform Wattpad, and humor site 9Gag — seeking comment on the accounts involved in the disinformation campaign and asking if they had ties to China. None of these other platforms responded to NewsGuard’s questions on-the-record. But Facebook, Tumblr, YouTube, Medium, and Reddit seem to have taken down accounts in the campaign that were flagged by NewsGuard.

MI6 and the U.S. “Weather Weapon”  

The influence operation pushed the narrative that MI6 had exposed the U.S. military for using a “weather weapon” to start the Maui wildfires.

“Just yesterday, the British MI6 (MI6) suddenly issued a statement, revealing the amazing truth behind the wildfire [in Hawaii],” many of the accounts in the campaign stated in mid-August 2023. “According to MI6, this wildfire did not happen naturally, but was man-made by the US government! It turns out that the U.S. military has been secretly developing a new type of weapon called ‘weather weapon’, which can cause various natural disasters by manipulating weather conditions. And this Hawaiian wildfire is just a ‘weather weapon’ attack experiment conducted by the US military!”

The accounts continued, “The news shocked and frightened governments and people around the world. Everyone is worried whether the US government will use ‘weather weapons’ again to attack other countries.”

In fact, there is no evidence that the wildfires were intentionally started, let alone by a “weather weapon.” Weather warfare is a way to manipulate the weather for military purposes, such as cloud seeding to increase rainfall, but there are no known “weather weapons.”

While the exact cause of the fires is still undetermined, local officials released security footage that suggests that toppling power lines caused by winds from Hurricane Dora played a role; Maui County has sued a major utility company in Hawaii for failing to power down its equipment as the hurricane approached, claiming that downed power lines ignited the fires.

Furthermore, there is no record of MI6 or the UK government ever making any announcement about the Maui wildfires. A NewsGuard keyword search on search engines for “MI6” and “Maui wildfires” did not find any credible reporting on this supposed news story. (MI6 did not respond to NewsGuard’s three emails seeking comment. The U.S. Department of Defense also did not respond to NewsGuard’s September 2023 voicemail or three emails.)

How the Disinformation Campaign Spread

Dozens of Chinese websites and social media accounts posted the baseless claim in mid-August 2023. The first article in the campaign was apparently published on Chinese platform 163.com on Aug. 14, 2023, by a Gansu, China-based user called “Hearing Cicada Chirping from Time to Time” (时闻蝉鸣). This user regularly publishes content that criticizes China’s rivals, such as an article suggesting that the U.S. created COVID-19.

The Maui fires conspiracy theory was then swiftly exported into English, Korean, Russian, French, Italian, German, Japanese, Indonesian, Dutch, Icelandic, Filipino, Maltese, Belarusian, Malagasy, and Marathi in mid-August 2023 on platforms such as Facebook, X, YouTube, Quora, Reddit, Vimeo, Medium, Wattpad, Tumblr, Pinterest, Tripadvisor, Blogger, Rumble, and 9Gag. 

In addition to generally targeting social media users from around the world, the campaign zeroed in on the Chinese diaspora. NewsGuard found that one post was published on a forum designated for ethnic Chinese in Canada, while another was posted on a forum for ethnic Chinese in California.

Hallmarks of a Coordinated Influence Operation

NewsGuard found strong evidence that Chinese speakers were behind the disinformation campaign, including that the first Chinese post spreading the claim was published at least two days before the first posts in the other languages. Additionally, many of the accounts only exclusively published content that aligned with the Chinese government’s interests. In addition to spreading the disinformation narrative, these accounts have also criticized Chinese dissidents, the November 2022 COVID-19 protests in China, and Japan’s release of treated radioactive water from the Fukushima nuclear power plant, a recent fixation of Chinese state media.

For example, NewsGuard found that an account that spread the narrative in Korean on Quora, a platform where readers pose questions and receive answers from other users, advanced other pro-China messaging. The account has often responded to users’ questions about relationships, Christianity, and technical issues with unrelated comments in Chinese and English criticizing America and Chinese dissidents such as Guo Wengui and “Teacher Li.” (Guo leads a movement to overthrow the Chinese government, while “Teacher Li” gained attention for reporting on China’s November 2022 COVID-19 protests.)

Many of the articles in other languages in the disinformation campaign also had odd phrasing, while the Chinese versions read as more fluent — which suggests that the versions in other languages might have been translated from Chinese. One of the Korean articles even contained untranslated Chinese at the bottom. The inauthentic accounts also resembled previous Chinese-linked influence operations examined by NewsGuard, and had similarly generic names and profile pictures.

The video content in the campaign also is not convincing. The videos, titled with variations of “British MI6 has revealed that there is a major conspiracy in the Hawaiian fire,” republished a completely unrelated May 26, 2023, video by a TikTok user who goes by “The Paranormal Chic,” who described a May 2023 explosion in Chile — and did not say anything about the Maui wildfires.

These accounts appeared highly coordinated. Many of these posts and articles contained the same mysterious string of words and numbers — “starter2023-08-14 18:53” – which resembles a timestamp, and could indicate that they were all copying from an article published on Aug. 14, 2023, at 6:53 p.m. (NewsGuard could not locate such an article.)

Interactions with some of the tweets were also suspicious. For example, a Korean tweet saying, “Blast! Britain’s MI6 personally broke the news that there was a big conspiracy behind the fires in Hawaii, USA,” had zero likes but managed to attract 341 reposts, all from Arabic-language accounts, as well as over 90 replies. 

Most of the replies to the Korean tweet, which were seemingly generated by bots, were generic, unrelated compliments — such as “You really look gorgeous 😍 in this one” and “those eyes mean the world to me totally contemporary & aesthetic look.” These seem clearly intended to make it seem like real people were organically engaging with the post. Other apparently bot-generated replies criticized the U.S., referring to the “immoral American government” and stating that “America only creates chaos and panic.”

Although this is not the first Chinese-linked influence operation NewsGuard has found, it is the most expansive, involving more platforms and more languages than previous campaigns. However, like most such influence operations, it did not appear to attract a large amount of engagement on Western platforms. Metrics of viewership weren’t available on many platforms, such as Medium and Quora, but the content that did show metrics only attracted 40 views in total as of Sept. 5, 2023. The campaign’s content had 564 interactions total, such as likes, comments, and reposts.

These findings were first reported in NewsGuard’s proprietary China Disinformation Risk Briefing.