Last Christmas Eve, NewsBreak, a free app with roots in China that is the most downloaded news app in the United States, published an alarming piece about a small town shooting. It was headlined “Christmas Day Tragedy Strikes Bridgeton, New Jersey Amid Rising Gun Violence in Small Towns.”
The problem was, no such shooting took place. The Bridgeton, New Jersey police department posted a statement on Facebook on December 27 dismissing the article - produced using AI technology - as “entirely false”.
“Nothing even similar to this story occurred on or around Christmas, or even in recent memory for the area they described,” the post said. “It seems this ‘news’ outlet’s AI writes fiction they have no problem publishing to readers.”
NewsBreak, which is headquartered in Mountain View, California and has offices in Beijing and Shanghai, told Reuters it removed the article on December 28, four days after publication.
The company said “the inaccurate information originated from the content source,” and provided a link to the website, adding: “When NewsBreak identifies any inaccurate content or any violation of our community standards, we take prompt action to remove that content.”
The operators of the website, findplace.xyz, did not respond to a request from Reuters for comment. The police declined to provide further comment.
As local news outlets across America have shuttered in recent years, NewsBreak has filled the void.
Billing itself as “the go-to source for all things local,” Newsbreak says it has over 50 million monthly users. It publishes licensed content from major media outlets, including Reuters, Fox, AP and CNN as well as some information obtained by scraping the internet for local news or press releases which it rewrites with the help of AI. It is only available in the U.S.
But in at least 40 instances since 2021, the app’s use of AI tools affected the communities it strives to serve, with Newsbreak publishing erroneous stories; creating 10 stories from local news sites under fictitious bylines; and lifting content from its competitors, according to a Reuters review of previously unreported court documents related to copyright infringement, cease-and-desist emails and a 2022 company memo registering concerns about “AI-generated stories.”
Reuters spoke to seven former NewsBreak employees, including five who said most of the engineering work behind the app’s algorithm is carried out in its China-based offices. The former employees requested anonymity, citing confidentiality agreements with NewsBreak.
Two local community programmes assisting disadvantaged people told Reuters they were impacted by erroneous stories produced by NewsBreak’s AI.
On three occasions in January, February and March, Food to Power, a Colorado-based food bank, said it had to turn people away because NewsBreak stated incorrect times of food distributions. The charity complained to NewsBreak in a January 30 email to NewsBreak’s general customer support email address, which Reuters has reviewed. The charity said it received no response.
Harvest912, a charity in Erie, Pennsylvania, emailed NewsBreak about two inaccurate, AI-based news stories which said it was holding a 24-hour foot-care clinic for homeless people, asking the outlet to “cease and desist” erroneous coverage.
“You are doing HARM by publishing this misinformation - homeless people will walk to these venues to attend a clinic that is not happening,” Harvest912 told NewsBreak, in a January 12 email seen by Reuters.
In response to Reuters’ questions, NewsBreak said it removed all five articles about the charities after learning they were erroneous and that the articles were based on incorrect information on some of the charities’ web pages.
Without providing a reason to Reuters, NewsBreak added a disclaimer to its homepage in early March, warning that its content “may not always be error-free”. Newsbreak generates revenue by showing ads to its users, who are predominantly female, above the age of 45, without college degrees, and live in suburban or rural parts of the U.S., according to the seven former employees and a 2021 company presentation reviewed by Reuters.
The company launched in the U.S. in 2015 as a subsidiary of Yidian, a Chinese news aggregation app. Both companies were founded by Jeff Zheng, the CEO of Newsbreak, and the companies share a U.S. patent registered in 2015 for an “Interest Engine” algorithm, which recommends news content based on a user’s interests and location.
NewsBreak told Reuters that the patent was assigned by Zheng to both companies because “some of the concepts were developed from Jeff’s time at Yidian” and that NewsBreak is “U.S.-based” and “U.S.-invested”. The shared patent has “absolutely no bearing on the company and its operations”, NewsBreak said in written responses to Reuters, describing the technology referenced in the patent as “outdated”.
Company memo
A May 2022 company memo from a NewsBreak consultant to Zheng, reviewed by Reuters, raised concerns about NewsBreak’s use of AI tools to re-publish stories from local news sites under five fictitious bylines.
“I cannot think of a faster way to destroy the NewsBreak brand,” Norm Pearlstine, former Executive Editor at the Wall Street Journal and the Los Angeles Times who was working at the time as a consultant to NewsBreak, wrote in the memo to Zheng.
In an interview after NewsBreak gave him permission to speak with Reuters, Pearlstine said he learned of the practice from a NewsBreak colleague.
“I question the legality of creating fake accounts using content publishers put behind their paywalls. If I had learned about the practice while at the LA Times, I would have instructed our lawyer to seek a restraining order and sue for damages,” wrote Pearlstine, whose six-month consulting role at NewsBreak in 2022 consisted of advising the company about U.S. editorial businesses.
Pearlstine, who confirmed the memo was authentic, attributed the lapse to a lack of journalistic experience. “A fair number of people on the staff were either new to journalism or new to the U.S. market. That was part of the reason I felt I had to be very direct and very explicit in explaining why I thought this was important,” he told Reuters.
NewsBreak said the news stories referenced in Pearlstine’s memo were a “limited experiment in three U.S. counties” to aggregate third-party content, and that the effort was disbanded after producing ten articles. The company denied going behind paywalls and said it used “snippets” of articles that were publicly visible to produce complete news stories using OpenAI.
NewsBreak also pointed Reuters towards Zheng’s emailed response to Pearlstine, saying he recognized the problem and asked his team to fix it.
OpenAI told Reuters its policies prohibited using its technology to mislead people. In 2022, Patch Media, which operates digital local news feeds in every U.S. state, reached a $1.75 million settlement in a lawsuit against NewsBreak for copyright infringement, according to court documents reviewed by Reuters, which alleged that NewsBreak republished Patch’s news stories without permission or credit. Patch did not respond to a request for comment. NewsBreak said the settlement was not an admission of wrongdoing.
Emmerich Newspapers, which operates newspapers in Mississippi, Arkansas and Louisiana, reached a 2021 settlement with NewsBreak in a lawsuit alleging copyright infringement related to NewsBreak’s use of Emmerich’s content without permission. NewsBreak said the settlement was “amicable.”
Another copyright lawsuit is ongoing. The two parties are “embroiled in additional lawsuits which we are vigorously defending against,” NewsBreak said.
Wyatt Emmerich, the company’s president, said the lawsuit against NewsBreak involved “verbatim copying of content”. He added: “What worries me in the future is that news aggregators could use artificial intelligence to slightly rewrite our stories which would make proving copyright infringement much more difficult. I have witnessed instances of this happening already on news aggregation sites.”
China roots
NewsBreak is a privately held start-up, whose primary backers are private equity firms San Francisco-based Francisco Partners, and Beijing-based IDG Capital, NewsBreak told Reuters.
NewsBreak presents itself as U.S.-based and U.S. invested, but was initially created as the subsidiary of a popular Chinese news aggregation app which was part-owned by a Chinese state-linked media entity.
Francisco Partners declined to answer questions about its investment in NewsBreak. IDG did not respond to repeated emailed requests for comment.
In February, IDG Capital was added to a list of dozens of Chinese companies the Pentagon said were allegedly working with Beijing’s military. IDG Capital told Bloomberg in February that it has no association with the Chinese military and does not belong on that list. NewsBreak did not comment on the finding.
Yidian, the Chinese aggregation company, divested from NewsBreak in 2019 because “its management team at the time did not understand the U.S. market”, Zheng said. Until then, Li Ya, the president of Phoenix New Media, a Chinese state-linked media firm which held a 46.9% stake in Yidian, had been a director at NewsBreak, according to corporate records.
Yidian continued to describe NewsBreak as its U.S. version on its website until 2021, according toThe Wire China.
Yidian in 2017 received praise from ruling Communist Party officials for its efficiency in disseminating government propaganda. Reuters found no evidence that NewsBreak censored or produced news that was favourable to the Chinese government.
A NewsBreak spokesperson said there was no ongoing commercial relationship with Yidian. Yidian, Phoenix New Media and Li Ya did not respond to requests from Reuters for comment.
About half of NewsBreak’s 200 employees are China-based where they are engaged in R&D, the company said.
A 2022 company roster reviewed by Reuters showed that 100 of NewsBreak’s 137 engineers at the time were based in China.
Five of the former NewsBreak employees said CEO Zheng divides his time between China and the United States.
Zheng, who was born in China, is a permanent resident of the United States and his family relocated to the U.S. early last year, the company said.
Reuters found five job advertisements NewsBreak posted on Chinese job sites seeking data analysts or engineers for its Beijing and Shanghai-based offices capable of “in-depth mining” of “massive user behaviour data” from the app’s U.S. users.
A Republican aide to the U.S. House of Representatives Foreign Affairs Committee told Reuters the use of Chinese-based engineers by Newsbreak raised possible concerns that American user data can be accessed in China. The aide declined to be identified because they were not authorised to speak to the media.
In a recent high-profile case, U.S. officials warned that TikTok, whose parent company is the Chinese firm ByteDance, could be compelled by the Chinese government to use its algorithm to control what kind of news is viewed by Americans and hand over their data.
TikTok, the most downloaded short video app globally, with 170 million U.S. users, now faces a forced sale or a U.S. ban.
In response to Reuters questions, TikTok said it was planning to offer third parties more access to examine its code and verify the app functions as intended. Zheng told Reuters that NewsBreak complies with U.S. data and privacy laws and is maintained on U.S.-based Amazon (AWS) servers. “Staff in China only access anonymous data stored on AWS servers in the U.S.,” he said. Amazon declined to comment.
NewsBreak also said that as a U.S.-based business it was not subjected to Chinese data laws.
Pearlstine, the former NewsBreak consultant, said NewsBreak’s ability to demonstrate it is a U.S. company was critical.
“The long term health of NewsBreak was dependent on its being perceived as a California company and that the more the leadership was in Mountain View, the better it would be for the company,” he said.
And absolutely no one was shocked.
Wew at least the app wasn’t corrupting the youth by revealing an ongoing genocide.
The article raises some great issues, but the weird xenophobia is a massive blight on an otherwise good article.
Where is the xenophobia?
It regularly uses descriptors like “Chinese roots” and “engineers based/born in China” and just leaves them hanging. On it’s own, being from China is superfluous information…unless the reader has a bias against things from China. Ties to the Chinese govt/CCP is a separate matter entirely, and if that’s the connection they’re making they should just say that.
It would be no different from talking about Isreal’s war crimes, but instead of saying “X party is affiliated with Isreal’s govt” you just said “jews”.
“Chinese roots” was a sub-heading. Did you read the paragraphs underneath? As for why it matters that the application and its employees are from China - pay closer attention to international relations between the US and China. The US banned Huawei because it was concerned about the influence of the CCP on the company and is in the process of banning an extremely popular mobile application currently (TikTok) because it has links to China. The US and China are major foes within the Pacific region: look at America’s behaviour around the issues of Taiwan and AUKUS. That is the important context you seem to have overlooked here.
The most popular news application in the US having links to China is highly relevant when the two nations see each other as antagonists on the global stage, when the US has a recent history of banning popular Chinese products and services because of a purported or real link to the CCP, and when the services themselves are being used to spread misinformation and, potentially, disinformation.
Yes, the examples I cited were used multiple times. And yes, I am aware of national news in the US concerning China, it was not overlooked context.
The relevant context you might be overlooking are the Red Scare and Japanese internment. It’s not the first time the US has had the looming threat of international espionage, and it’s also not the first time that using a person’s/company’s nationality to infer their true intentions was misguided.
It is not an outrageous thing to say: being Chinese does not make you a CCP operative.
The most popular news application in the US having links to China is highly relevant…when the services themselves are being used to spread misinformation and, potentially, disinformation.
To quote the article,
Reuters found no evidence that NewsBreak censored or produced news that was favourable to the Chinese government.
To echo the sentiments above, most of the article is great journalism. The Chinese job listings, the former connection to Yidian, their use of AI and statements from Pearlstine. All important info to see reported. I just wish they would stop saying “China” or “Chinese” as shorthand for “CCP”. All I can hear is Trump annunciating CHYAI-NA.
I wish the best for the Chinese. I want Chinese people to feel safe when living abroad, without their govt breathing down their neck. I want China to prosper ethically and sustainably.
I think you’re imparting your own biases on the article here. The reporting on the “Chinese roots” of the company is provided as context for the reader because, as I explained above, China’s alleged influence over American society and politics is very relevant at the moment - regardless of whether you agree that it is happening. As far as I know, there was also zero evidence provided by the US government of CCP-guided interference by Huawei or TikTok, yet it moved ahead with bans anyway. Particularly within the context of the reasoning for banning TikTok, NewsBreak’s connections to China are extremely relevant. Note, in particular, this line from the article:
In February, IDG Capital was added to a list of dozens of Chinese companies the Pentagon said were allegedly working with Beijing’s military.
You accuse the outlet/journalist/editor of xenophobia and quote their findings of zero links to the CCP as evidence of this, but I would argue that this is actually evidence of the complete opposite. They are doing their best to provide a balanced view of the situation by reporting the lack of connections in contrast to the allegations or implications. If they actually had a genuine xenophobic intention with this reporting, there would be absolutely zero reason to include that line.
It is not an outrageous thing to say: being Chinese does not make you a CCP operative.
The article doesn’t state or imply that, though. What it does do is provide information to the reader that is, again, relevant because of the current situation in US politics where you do have politicians alleging or implying that being Chinese, even ethnically, equates to being a CCP operative.
I just wish they would stop saying “China” or “Chinese” as shorthand for “CCP”.
This is standard in the field of international relations, though. We always use the name of the country or the capital city of the country when referring to it on the international stage. You are suggesting that we should refer to China as “the CCP”, the USA as “the Democrats”, Australia as “Labor”, etc. That is obviously an extremely confusing and illogical way of presenting information to a global audience.
I think the fact that you are so concerned about how Trump might hijack this type of balanced reporting is a massive tell that you are not reading it impartially. You are approaching it from the position of “the West has an anti-China bias” (a position I don’t necessarily disagree with) and are then trying to link the dots between the information provided in the article and your world view. You need to understand that reporting facts, that may be parroted in bad faith by political groups and their supporters to reach an inaccurate/unsupported conclusion, is not the same as an opinion piece where the objective is to emotionally manipulate the reader or lead them to a particular conclusion. Reuters has an extremely good reputation for highly factual and unbiased reporting and there is nothing in this report to suggest otherwise.
there was also zero evidence provided by the US government of CCP-guided interference by Huawei or TikTok
There were a couple decades of direct evidence that Huawei was cooperating with the CCP.
Disinformation on TikTok and data collection via the app are both well documented, but doesn’t seem any different from other social media platforms. The TikTok ban seems mostly politically motivated to me. I suspect whatever happens, their lawyers will fight it and win, until the govt finally cracks down on data collection in general, and we finally get something like GDPR.
You are suggesting that we should refer to China as “the CCP”, the USA as “the Democrats”, Australia as “Labor”, etc.
No, I’m literally saying the opposite. We should refer to the CCP as the CCP, and China as China. My last paragraph was intended to highlight the difference, but perhaps you didn’t read that far.
The rest of your comment is ad hominem, which doesn’t interest me.
There is not a single attack on your character at all in that reply. If you don’t want to engage, fair enough. Just be honest about it instead of trying to smear others so you look like you “won” by default due to some imagined moral high ground.
What kind of madman uses an app to get their news?
If you’re getting your news via your smartphone or pc, you’re using an app
Most people… I mean, the one I use is just an RSS feed aggregator. But in the age of smartphones, people just get an app from their favorite sources and never check others it seems.
