Manipulation Center
One morning in 2022, the Brazilian executives of Kwai, a social network for short videos, arrived at their headquarters in São Paulo. The first item on the agenda: discussing the memos sent to them by their head office in Beijing, located eleven hours ahead of Brazil. That day’s six-page memo, written in English and Mandarin, had an almost bureaucratic title, “Risks with KwaiCuts and the 2022 Brazil Election,” but the subject itself was delicate and had been shrouded in secrecy. Its contents showed the dark side of the social network’s operations in Brazil.
The term “KwaiCut” refers to all the material Kwai commissions from content agencies. In the battle to expand their audience, Kwai purchases content that can attract user interest and increase the time spent on their platform. Sometimes Kwai will tell the agency what they are looking for. On December 29, 2022, for example, they commissioned videos about Pelé, who had recently died. Most of the time, however, Kwai will not make specific requests. Their only goal is to have viral content, regardless of whether it is fake or untrue.
That morning’s memo listed content generated by agencies over the past two weeks and also flagged 26 “problematic” videos about the upcoming Brazilian election. These videos made up 22% of all the platform’s KwaiCut videos with political content. The document did not elaborate on the content of the videos but simply stated, in broad strokes, that they contained “political misinformation.”
The company could take a series of energetic measures to block untruthful posts, but the memo itself raised a question: should the company exclude these videos and avoid being penalized and fined by Brazilian authorities? Or should the company deliberately ignore these false posts and use the content to attract a wider audience?
The executives understood what was at stake. They also knew about the irregular practices being adopted by the network. One excerpt of the memo, to which piauí had access, makes this abundantly clear: “If the accounts are found to be linked to the company, this could severely impact on our operations in Brazil.” AYep, nother section of the memo contained the following warning: “Brazil media outlets are doing more investigations into the role of social media apps of spreading election misinformation.”
The memo detailed five possible strategies, ranging from stopping the production of “political KwaiCuts until after the election” to “maintain[ing] status quo,” which, according to the document, would involve “tolerat[ing] all the risks highlighted” and dealing with “negative press, data requests or complaints/lawsuits on a case-by-case basis.” In the end, the company didn’t adopt one uniform strategy. Some content was banned, but generally speaking, disinformation ran rampant and unobstructed on the platform.
The document is a rare case in the underground world of Big Tech. Large social media companies across the globe—Facebook, Instagram, TikTok, X (formerly known as Twitter)—have been accused of spreading lies and hate speech, which are then turbocharged by the invisible hand of audience-seeking algorithms. But this is the first time there has been material proof that a social network directly commissioned viral content that arrived riddled with fake news.
“The fact that a platform is hiring an agency to produce content is itself shocking,” says researcher Rose Marie Santini, coordinator of NetLab at the Federal University of Rio de Janeiro. “These platforms claim they aren’t subject to Brazilian law because they aren’t technically media companies. They say that, as tech companies, they cannot be held accountable for their content. But if they’re hiring content agencies, then they are producing content, regardless of whether a third party is involved.”
For two months, piauí interviewed eight current and former Kwai employees, six of whom held management or directorial roles; all of them spoke on condition of anonymity, fearing retaliation from the company. Piauí also interviewed former third-party agencies and contractors connected to Kwai in Brazil. The picture these conversations paint—and the documents piauí accessed—reveals a deliberate policy that encouraged irregularities for the purpose of procuring viewership, be it by spreading fake news, deliberately cloning accounts on other platforms, or even boosting the profiles of presidential candidates, none of which social networks are legally permitted to do.
Kwai is a discreet but powerful social network. As of today, there are 48 million active users in Brazil, according to data obtained from the platform ComScore. That’s fewer than Instagram, Facebook, and TikTok but twice as much as X, which only has 24 million users in Brazil.
Kwai (pronounced “kwye,” though it became known in Brazil as “ka-wye”) landed in Brazil in June 2019. Back then, the app, which was founded by Kuaishou Technology, was already a giant in its native China. When it arrived here, it found a space for itself among the lower middle classes, especially in the North and Northeast. While TikTok, another Chinese platform, was best known for its dance routines, Kwai became popular for its mini-soaps and dramatizations of romantic scenarios, at times humorous and other times edifying. A large part of their users, however, were drawn to disturbing material shown to some degree on other platforms as well, for example, of car crashes complete with dead bodies and brutal violence against women.
Jair Bolsonaro was the first presidential candidate to pick up on Kwai’s potential. He created his profile on May 20, 2022, a few months before the election. In China, content teams were elated about his arrival on the platform. According to three people who worked for the company at the time, the head office immediately began promoting Bolsonaro’s posts. Brazil’s Superior Electoral Court regulations state that a campaign can boost its candidate’s content—that is, it can pay a social network to give the candidate greater visibility—but the network itself does not have this authority. Everything points to Kwai promoting Bolsonaro for commercial reasons, as the company wanted to increase viewership and engagement. There is no evidence that the two were ideologically aligned.
The team in Brazil had some concerns about the irregular measures adopted by the head office. Messages between two colleagues in the department responsible for political content suggest an air of apprehension. “PYMK got a lot of traction and China gave it three boosts at the end of the month,” read one message obtained by piauí. The colleague replies: “FFS. That’s insane.” PYMK (People You Might Know) is a feature that recommends profiles to users who sign up to the network, while “boosts” increase a profile’s visibility, allowing it to reach more people.
In July 2022, the Brazilian fact-checking agency Aos Fatos saw that Bolsonaro’s fake news posts were being promoted by Kwai. On their site, they posted: “Videos in which President Jair Bolsonaro makes false claims were viewed over 13 million times on Kwai [. . .]. This figure represents close to half (45%) of the president’s current reach on that platform. Posts containing disinformation shared by Bolsonaro were viewed on average three times more than other posts, for a total of 435,000 views each.” In one of these, Bolsonaro attacked Brazil’s electoral polls, claiming that public opinion institutes had failed to predict he would make it to the second round in the 2018 elections, an untrue statement.
That July, two months after Bolsonaro, Lula also launched an account on Kwai. His appearance on the platform was equally celebrated and promoted by the Chinese head office. But research conducted over the course of two days, twenty days prior to the election, showed that Lula and Bolsonaro did not receive equal treatment. The study, to which piauí had access, indicated that 46% of Lula’s growth came from recommendations to follow his profile (which were boosted), compared to Bolsonaro’s 82%. This gave the latter a significant advantage during the sample period and increased his digital presence in the Northeast, which is an especially important region as it contains both the bulk of Kwai’s viewers and of the Workers Party’s voter base. (Today, Lula has 6.8 million followers compared to Bolsonaro’s 5.9 million.)
Piauí consulted the Superior Electoral Court, which indicated that it had a good relationship with Kwai’s management during the electoral campaign. When asked, Kwai would promptly remove any untruthful content. In February 2022, in an effort to combat fake news in the election, the court put forward a cooperation agreement that would open a direct channel to Big Tech and streamline the takedown of fake content from their platforms. Kwai signed the document. And yet, behind closed doors, their manipulation center remained active.
In April 2022, election year in Brazil, Tecnoblog, which specializes in news from the digital market, published a piece about the glut of untruthful content on Kwai. The title of the post was “Kwai is promoting fake news about electronic ballots and Covid-19 vaccines.” But what they didn’t know at the time was that things were actually worse: the false content was part of a packet commissioned by Kwai. This information can be found in an internal memo scrutinizing the article that was circulated to Kwai employees the following month. In it, Kwai expresses concern about the report and states at one point that “the videos [containing fake news] were published by KwaiCut accounts.” In other words: they were part of the content Kwai commissioned from agencies.
On October 1, 2022, the eve of the first round of the Brazilian general election, Kwai uploaded a video at 4:41 p.m. showing the CNN 360° studios and journalist Daniela Lima, at the time a host for CNN Brazil, under a fallacious banner: “Armed forces could intervene at any moment!” The fake post got at least 1.5 million views. Piauí was able to verify that the field designating the source of the video—information that only appears in the platform’s internal systems—read “SC/KwaiCut.” In other words: this was just more material from Kwai’s arsenal of commissioned content.
In the decisive days of the election, Kwai was overrun with fake news purchased by the company. And they got thousands of views. A few examples:
* October 2, the day of the first round of the election: A misleading post stated “Neymar backs Lula.” Over 600,000 views.
* October 3: The lie of the moment was that Bolsonaro had decided to leave the race. It broke the one-million-view mark.
* October 4: Lula was at risk of “losing his presidential candidacy.” More than 900,000 views.
* October 5: A video claimed that Lula was incentivizing robberies. Views surpassed 500,000.
Kwai hires content agencies and authorizes them to share materials directly on their platform. At the same time, evidence shows that the company itself, even when alerted to untruthful content, will turn a blind eye and officially allow the material to be shared. Piauí accessed a video that spread lies about “military intervention” and bore the digital stamp “allow.” Visible only to people with access to Kwai’s internal network, this stamp demonstrates that the video was evaluated by the content moderation team and allowed to remain online. According to employees, moderators were under pressure to keep videos online for the sake of views.
In the final stretch of the election, a moderator sent the following message through an internal system: “Hi everyone, another kc [KwaiCut] account made a post whose content violates election day rules.” The author, who also sent his note to company leaders—some of whom have Chinese names—advised that content moderators had determined the post should be deleted “for exhibiting extremist behavior.” The message goes on: “We’d be grateful if you could please reply to the last two cases we referred to you, as these accounts continue to impede our efforts to regulate electoral content.” In another message, a colleague backs up the original request: “We’d be grateful if KC could refrain from publishing any risky content on election day.” It’s unclear to piauí what came of these warnings. The author of the first message did not wish to be interviewed for fear of being identified.
Due to the high volume of posts—about 2 million videos a day—Kwai hired third-party contractors to bulk up the content moderation team. These new hires were identified in their contracts as telemarketers. Among them was Julio Cesar Nogueira, 23. For seven months, he spent six hours a day watching videos that turned his stomach, including some with explicit violence and child pornography. These he instantly blocked. But when it came to false content, he didn’t even bother. “Fake news was allowed,” he said. “Sometimes I’d block things, then they’d get sent to China [the team of moderators based in Beijing] and get the go-ahead. They only took down content when there were tons of complaints.”
Even though Nogueira did not keep any screenshots, he says he saw a piece of fake news that went so far as to accuse President Lula of being a “rapist” and a “murderer.” In fact, there was one video plastered with the words “LULA IS A MURDERER!” in block letters, accompanied by a piece about the death of late mayor Celso Daniel, which makes no such insinuation. In October, Nogueira and five colleagues were fired for shaking a vending machine that had swallowed their money and didn’t dispense their item. He sued Teleperformance Brasil, the third-party company Kwai had hired. They did not reply to requests for comment.
Kwai’s head office in Beijing pays content agencies directly. The platform makes no attempt to hide its relationship with these agencies, which they refer to as “partners.” But this is not the whole story. Piauí was able to access a spreadsheet corresponding to payments made in June 2022. Twenty-one agencies were listed on it. Kwai paid them via Payoneer, an American financial services company and PayPal competitor. The best-paid agency in that period was IMP Digital—formerly called Imperium—which received a total of $35,353 plus a $1,580 bonus.
I reached out to a few of these agencies under the guise of wanting a job with KwaiCut. At IMP Digital, I was directed to a producer of Agência Realeza, who would be responsible for training me. I’d learn how to download YouTube videos with the app Snaptube and edit them with CapCut. The basic idea was that the editors would download YouTube videos, pull out short clips, and put “imaginative” banners on them to help get users’ attention and increase engagement. When I asked if this violated copyright, the producer reassured me. As soon as I’d finished registering with KwaiCut, Agência Realeza would send me a foreign phone number. “Our account is attached to a country with copyright regulations that aren’t as strict as Brazil,” she explained. She did not identify the country in question.
Let’s Go, another agency featured on the spreadsheet, received $28,000 in that same period. They sent me a file with a list of the kinds of videos they wanted produced, including “weird news” and “fights.” I needed to post a minimum of 100 videos per month. Eventually, the agency would give me a phone number I could use to log on to Kwai. There was an alert in block letters: “REMEMBER THE COUNTRY CODE IS NOT BRAZIL BUT ZAMBIA +260.” This warning made it clear that much like IMP Digital and Realeza Agência, Let’s Go Agency also operated out of another country.
At all these agencies, payment is dependent on the number of monthly views for each video. According to the spreadsheet piauí was able to access, Mirah received $19,000, with figures ranging from $30 (for accounts that got between one and two million views) to $2,000 (for accounts with more than 50 million views). If an account had fewer than one million views in a month, they made nothing. At Blue Pink, which was paid $5,600 according to the spreadsheet, you received more points for posting videos about “gambling,” “sports,” and “kids and babies.” If a video featuring “kids and babies” or “humor” got half a million views, then the agency would charge double, as if it had actually received a million views. In the case of “sports” or “gambling” videos, the number was tripled. These are some of the topics that get the most views on the platform.
When approached by piauí for an interview, none of these agencies wanted to explain their business models. When I informed Let’s Go that I was a reporter with piauí, I received a response from a 20-year-old producer who confirmed he was authorized to speak on the company’s behalf. I asked why the agency was using a foreign phone number, and he replied: “The Zambia country code has always been a mystery to me. It’s China [the Kwai team in Beijing] that creates the accounts and sends us the number.” The day after the interview, the young producer told me he'd made a mistake and couldn’t actually speak on behalf of Let’s Go. (For this reason, piauí has decided to withhold his name.)
That same day, Letícia Brandão Silva, who owns the Let’s Go Agency, contacted piauí to also inform me that the young man had not been authorized to speak on their behalf. “He’s a contractor, not an employee. Each company we subcontract is responsible for their own conduct.” Brandão Silva, however, refused to take my questions about her agency’s activities. “Don’t use our company’s name. Ours is a labor of love, and we do what’s right. We bless lives,” she said.
In 2018, Facebook found itself embroiled in a major scandal. Two former employees of the British group SCL, then owner of the data firm Cambridge Analytica, revealed that the company had for two years been using an online questionnaire to harvest the data of 87 million Facebook users in the United States and map potential undecided voters. During the election campaign, these voters were bombarded with political ads in favor of Donald Trump, who ended up beating his Democratic opponent, Hillary Clinton, in the race for the White House.
Following this scandal, Mark Zuckerberg, founder and CEO of Facebook (now Meta), admitted the platform’s mistake but outsourced the blame when testifying before the US Senate: “When we learned in 2015 that Cambridge Analytica had bought data from an app developer on Facebook […], we did take action. We took down the app, and we demanded that both the app developer and Cambridge Analytica delete and stop using any data that they had. They told us that they did this. In retrospect, it was clearly a mistake to believe them,” Zuckerberg said. Brittany Kaiser, a former director at the data analytics firm and one of the whistleblowers in the case, told another version of the story. In the Netflix documentary The Great Hack, she claims Facebook knew the stolen data was being used for electoral purposes during the campaign. Facebook, with its more than three billion active users—40% of the world’s population—has had all sorts of accusations leveled against it. In one particularly horrendous case, the platform failed to control the spread of lies and hate messages in Myanmar, which contributed to the persecution of an Islamic minority in the country, resulting in 25,000 deaths and prompting 700,000 people to flee the country. In Brazil, Facebook allowed a gamut of rumors to spread, including the false claim that vaccines cause Zika, which had an impact on childhood immunization. Zuckerberg has always downplayed his company’s responsibility in these cases. Meanwhile, behind the scenes, Facebook’s radioactive content keeps user engagement at a rolling boil.
And yet not even Zuckerberg’s network has been accused of commissioning content agencies to deliver viral content awash with fake news. Nor has it been accused of creating counterfeit accounts, another Kwai irregularity that piauí uncovered in interviews and documents.
One afternoon in November, I went to an apartment in São Paulo for a look behind Kwai’s curtains. A company staff member opened the door on one condition: I couldn’t use any device with an Internet connection. She took me to her office at the back of the apartment, where the cameras on both her computers were taped so no images could be captured and put on ambient music in the room to prevent eavesdropping. “I’ve worked in tech for many years,” she said with a smile.
Using an internal Kwai system called Noah, she showed me a search filter that pulled up all the agency content. There were more than 100,000 KwaiCut videos, all catalogued by agency name, video category (such as TV series or movie), and a label that read “vertical pirating account,” which demonstrates that Kwai itself considers agency content to be “pirated.” But there was more: after applying another filter to the system, a list of content marked “CC-0” appeared, with the label “pirating account.”
The staff member explained that the CC-0s were user profiles cloned by Kwai from other social media networks, especially TikTok. In other words, someone will open an account on TikTok or another app, and then Kwai will clone this account and add it to its own platform, thereby increasing user volume. An enormous universe was displayed on the monitor: more than 3.5 million pirated accounts, both active and inactive, banned from the network due in part to user complaints. So, although Kwai may report that it has 48 million users in Brazil, the real number may be a bit lower than that.
Weeks later I was given access to a message—sent on November 19, 2021, at 2:12 a.m. Brasilia time—that corroborated what the Kwai staff member had told me about those CC-0 profiles. In the message, Kaiyun Qian, an international legal director based in Beijing, issued an urgent warning: Kwai had just been notified by the São Paulo State Supreme Court that it had cloned an account in the name of “Patricia Shiman [sic].” The court said the account had to be deleted and asked for the IP address and device location, in order to identify the person responsible for the fraud.
One minute later, Kaiyun Qian sent another message, likewise written in English: “The account was confirmed by TNS [moderation team] to be a CC-0 account. Therefore, we do not have any chance to win [the lawsuit].” The director then went on to talk about consequences and measures: “Not only we will bear the cost and damage but may be held criminal liable if we reject to furnish the evidence of the account information. The judgment will also be published and set as a precedent for future cases, unless we made a settlement to keep it confidential [sic].”
And this is exactly what happened. When approached, Patricia Shimano—the correct spelling—attested that she couldn’t discuss the case because she had signed a confidentiality agreement with Kwai. There were more cases like hers, as shown in an internal memo with a self-explanatory title: “On-going Litigation or Administrative complaint in Brazil – FAKE ACCOUNT.” The document, to which piauí also had access, gives key information on Shimano’s case and five others, including which court is handling each case, plaintiff’s name, type of complaint, and stage of the lawsuit. There is also a “status” field, which specifies the amount paid to each complainant. According to this document, Shimano received roughly $3,700 (19,925 reais) for her silence.
Another case involves the Goiânia-based company Sora e Nogra, which sells e-books about mother-in-law/daughter-in-law relationships. In a petition to the São Paulo court, Sora e Nogra alleges Kwai used a fake account to clone its content, presenting dozens of examples. The court gave Kwai five days to retrieve data from the fake accounts and identify the party responsible. The platform claimed this information had been deleted more than six months earlier. (Under Brazil’s Internet Bill of Rights, data cannot be stored for longer than six months.)
Sora e Nogra argued to no avail that Kwai was acting in bad faith. On July 31 of last year, Judge Rodrigo Galvão Medina, of the 9th Civil Court of the Central Forum, ruled in favor of the social network, and the Goiás company even had to cover court costs. However, the document obtained by piauí shows that Kwai didn’t play fair. The following information can be found in the “status” column of the Sora e Nogra case: “Collecting information now; amount not available; no penalty for not providing information; CC-0 account.” In other words, Kwai’s internal system indicates that the account had effectively been cloned from Sora e Nogra’s profile.
Despite the many lawsuits, Kwai continues to clone accounts. Another victim of this is attorney João Raposo, whose TikTok page has more than 200,000 followers. At piauí’s request, Raposo conducted an experiment. On November 25, he posted a video on his regular TikTok account to see if the content would be pulled into his fake account on Kwai. An hour and a half after it was posted, the video had been replicated on Kwai. “It leaves you feeling helpless,” says Raposo, who plans to go to court. “I’ve reported it to the platform twice and nothing happens.”
A Brazilian executive who held a senior leadership position at Kwai has filed a lawsuit in the São Paulo Regional Labor Tribunal. Piauí was able to access the contents of this court case, despite its confidentiality, and has chosen to withhold the plaintiff’s name. The first document in the lawsuit describes how the platform operates and details its relationship with content agencies and cloned accounts. At one point, the text reads: “Kwai operated (and operates) a veritable factory of fake accounts with PIRATED content. […] Numerous content producers have complained of finding cloned accounts, believing this to be the practice of hackers or digital criminals. THEY NEVER IMAGINED that KWAI itself was the author of these accounts.”
The lawsuit claims that Kwai paid for false content and ignored moderator warnings; it also describes the former executive’s astonishment when she learned of the platform’s inner workings. “Contrary to what she [the former executive] expected from a company listed on the stock exchange, with an annual global revenue of around $13 billion and whose governance should be excellent, the company’s foreign leadership [the Chinese main office] not only ignored and rejected the complainant’s warnings but also labeled her as ‘crazy’.”
One case cited in the lawsuit hints at an air of secrecy within the company. The plaintiff elaborates on how Ian Lee, the Beijing-based director responsible for operations in Brazil, humiliated her at a meeting; she later discovered that Ian Lee wasn’t his real name. In a message thread obtained by piauí, a person working at the company asked “Ian Lee” for his LinkedIn name, because they wanted to add him and couldn’t find his profile. “You didn’t find me because I’m using a pseudonym,” the director replied. “Lee is not my real name. My former company [he later said this was TikTok] forced me to sign a clause that prevents me from working for competitors. I can reveal my real name and live as an honest person again after next summer.” Piauí was unable to uncover Ian Lee’s true identity.
Claudine Bayma, the general director of Kwai in Brazil, declined a request for an interview. Piauí then sent the platform’s press office a list of 24 written questions inquiring about the practices of posting knowingly false content, creating pirated accounts, and improperly boosting the visibility of presidential candidate profiles.
The social network responded with the following note:
Kwai has received no evidence of the information raised by the reporters and does not comment on lawsuits or court decisions. That said, it should be noted that Kwai prohibits the creation and publication of inappropriate content.
The platform uses 24-hour-a-day security mechanisms combined with AI and human analysis to identify and remove content that violates and infringes community policies as quickly as possible. Kwai also offers their users reporting channels where they can report inappropriate behavior on the app.
Kwai, an app for creating and sharing short videos, was launched in Brazil in 2019 and now has more than 48 million active users in the country, according to data from ComScore 2023.
Kwai is committed to building a respectful, healthy, and harmonious community that complies with its Terms of Service. The app’s personalized recommendation engine gives users access to a diverse range of content creators and includes suggestions of videos tailored to the user’s profile.
Protecting freedom of expression is one of Kwai’s principles. That said, the platform does not tolerate content that has the potential to spread false, misleading, or manipulated information or information that may otherwise be harmful to individuals or institutions. By ‘disinformation’, the app means both false information spread without harmful intent as well as the intentional dissemination of information known to be false by the user who posts or shares the content. To assist its users, Kwai has a Community Guidelines page.
Since 2022, Kwai has had a partnership with the fact-checking agency Aos Fatos, which helps identify material that could potentially be banned. Content can be removed from the platform or labeled as disinformation. Unverified videos are not given as much visibility, and repeat violations can lead to accounts being banned. Likewise in 2022, Kwai signed a memorandum of understanding with the Superior Electoral Court aimed at coordinating efforts to combat the spread of disinformation during the electoral process.
That same year, the app launched the first edition of its Transparency Report, marking another step toward ensuring an open exchange of information through ongoing transparency. Further underscoring the importance of accountability, Kwai published its fourth iteration of the Transparency Report in 2023. The document highlights the company’s actions and efforts to protect the integrity of the platform and its users, for example, by removing over five million videos that violated its policies and terms of use. In the first half of that year alone, it also banned almost 700,000 accounts that violated its Community Guidelines, either on a temporary or permanent basis. During that same period, Kwai received 73 requests from official authorities to restrict content and 1,202 reports of intellectual property violations. All were complied with.
As the note made no mention of the content agencies hired by Kwai, pirated accounts, or its track record of boosting presidential accounts, piauí resent its questions. The company did not provide any new information. When consulted by piauí, the fact-checking agency Aos Fatos said its work for Kwai “consists of verifying videos forwarded by the platform’s moderation team.” It also stated: “We do not take part in moderation decisions. In other words, Kwai decides whether a given post will stay up or go down.” Aos Fatos provides the same service to Meta and Telegram.
Kuaishou Technology, Kwai’s parent company, was founded in 2011. The following year, they entered the short video market and went on to become the second largest network of its kind in China, behind Douyin, which is the Chinese name of TikTok. Residents of rural China are Kuaishou’s target audience.
Between the late 2010s and early 2020s, the company began expanding into other countries, reaching Brazil and Indonesia, today the app’s second and third largest markets, after China (in Indonesia and India, the company is known as SnackVideo). But international expansion only took off in 2021, when the company went public on the Hong Kong Stock Exchange and raised $5 billion.
Before that, in May 2020, Kuaishou Technology tried to gain a foothold in the US market. Instead of calling itself Kwai, the app adopted the name Zynn, but its aggressive practices soon caught the attention of the US press. On June 9, 2020, Wired magazine released an article accusing Zynn of publishing content stolen from other platforms—just as it has been doing in Brazil. “Zynn, the Hot New Video App, Is Full of Stolen Content” the headline read.
After the report came out, the Apple and Google stores stopped offering the Chinese company’s app to users in the US, although it can still be downloaded from both stores in Brazil. Its public image tarnished by the report on its fake accounts, Zynn grappled with low demand and gave up on the US, closing its doors in August 2021, a little more than a year after its debut.
It’s still full steam ahead with the company’s marketing initiatives in Brazil. In 2021, Kwai signed a partnership with Band to launch a contest called Microfone Aberto—Open Mic—where participants compete for the chance to be a commentator on Jogo aberto, a daily evening sports round table. In the second half of 2023, Kwai’s logo was emblazoned across the soccer jerseys of Fortaleza, the top-flight Series A team in the Brazil Cup, which reached the Copa Sudamericana Final. Kwai is also now the official sponsor of A fazenda, a reality show on the Record television network. Last October, Globo TV announced that Kwai would be sponsoring its series Big Brother Brasil.