Q&A: Viola Zhou on the Challenge of Covering Chinese Tech

Sign up for The Media Today, CJR’s daily newsletter.

In January, DeepSeek, a Chinese AI company, shook Silicon Valley when it released a chatbot tool that matched the abilities of recent models made by OpenAI, a leading US firm, but for a fraction of the cost and hardware. DeepSeek’s innovations came as a surprise to tech firms that did not expect China to be a serious competitor in AI. Many of the responses—such as a post from the venture capitalist Marc Andreessen calling DeepSeek “AI’s Sputnik moment”—framed the news as a worrying sign of increased competition from China’s tech industry. Much of the media coverage did likewise, refracting the story through a geopolitical lens. 

This combative stance matches the outlook of the Trump administration. Last month, Vice President J.D. Vance gave a speech in Paris in which he staked a claim for the US on AI innovation, casting China as an authoritarian competitor to be battled. The Trump administration has reversed AI regulations put in place by the Biden administration in an effort to accelerate the industry, and also laid ground for a trade war with China, imposing 20 percent tariffs on goods from the country. The White House is now reported to be making a deal with Oracle, a US software company, to run the Chinese-owned video app TikTok in the US, after Trump paused a bipartisan Biden-era law dictating that it be sold or face a ban.

How the Chinese tech industry is thinking about these questions, however, is less immediately visible to US news consumers. To get some perspective, I asked Viola Zhou, a senior reporter who covers China’s tech industry for Rest of World, about the approach of company leaders, workers, and researchers in China, and the challenges of covering them from abroad. Zhou’s work often complicates the political narrative of competition and conflict, as was the case with her recent story on a Chinese electrical vehicle battery company named Gotion that has faced local opposition to opening a plant in Michigan. Our conversation has been edited for length and clarity.

CB: How would you describe the primary challenges of reporting on the Chinese tech industry now?

VZ: I sometimes joke that Chinese people don’t talk. In the case of an American company’s success, the CEO will be on TV the next day or on social media talking to everyone and collecting millions of followers. But in a Chinese company, this guy actually has to lay low, because attention is bad for a Chinese tech entrepreneur. Especially when it comes to English-language media, people are really nervous that whatever they say will be put into a political framework. I understand that fear; whatever they say becomes part of a US-China story. They are also reluctant to talk because in China, most of the mainstream media is run by the state. If you’re talking to domestic Chinese outlets, what you say won’t end up sounding critical of the authorities. But when you’re talking to international media, you have no control over that framing. A tech entrepreneur who is doing business in China doesn’t want to talk to international media because they can’t control if it’s going to be a positive or a negative story. So there’s a lot of reluctance, fear, and nervousness about talking to the media. 

Another challenge is just access. It’s very hard to get a journalism visa in China if you work for a foreign outlet. And Chinese citizens are not usually allowed to work as foreign correspondents based inside China. So there are very few journalists who are able to report on the ground. Most journalists, like me, have to do remote reporting on China by stalking people on the internet or doing phone calls. 

A recent story that you wrote about DeepSeek focused on how the company is convincing top AI talent in China to stay there rather than go work for a US company. One of your sources told you, “America thinks China’s trying to unseat America, but the truth is that young people were inspired by new technology developments such as OpenAI.” So while the US government and Silicon Valley largely view the Chinese and American tech industries as in competition with each other, would you say that Chinese tech workers don’t necessarily see it that way?

A very common stereotype in mainstream media coverage of Chinese tech is the idea that every business, and even every individual, does things to compete with America; that everyone in China has political motivations. But actually, people do things for all kinds of reasons. For example, I was looking into the DeepSeek researchers after their recent model was released. A lot of the researchers got super active on X. They were very proud of DeepSeek. They were interacting with researchers at American companies to discuss AI technical problems that I couldn’t understand. Just seeing that exchange was refreshing. If we read the headlines, it’s all about how China came up with something that competes with America or is better than America, or how the US ban on exporting advanced semiconductor chips to China led China to invent an AI model that doesn’t need so many chips. But when I was reading the X posts from the actual employees, I got a sense that they considered it more like a scientific research breakthrough. They came up with something that they’re really proud of and are eager to share with the rest of the open-source community. They wanted to show they have the ability to do something new. And they wanted their peers, whether in America or other countries, to see that. That’s their motivation. 

Maybe there’s some sense of nationalism. But when I talk to people in the industry, they want to build their personal brands. Companies want to build the company brand. So they have motivations that are more about business, about branding, about reputation, about pride. I’m not saying politics is not relevant. But when we approach these stories, when we are talking to these people and trying to find out their personal experiences, it’s important not to go into everything with a framing of US-China competition. There’s more cooperation than people think, especially in the research open-source community. Chinese researchers follow developments in America very closely. Not to be like, Oh, America has something. We have to come up with something so China can win. They get inspiration, and then they invent something new to make their own business better and to survive the competition in the Chinese tech industry, which has become very saturated. 

If we are reading about the US-China tech war, it seems as if every business wants to be part of it. But in reality, the Chinese tech entrepreneurs I report on really want to get out of the war. They are eager to do business in America. In the case of TikTok, they wanted to get American users. But they are running into this tech war. It causes them a lot of political and regulatory trouble, and then they have to deal with scrutiny or suspicion. From the Chinese companies’ perspective, they actually want to get out of politics. They have to navigate a lot of regulations at home, and then they have to navigate the US-China geopolitical tension when they go abroad.

How would you characterize the Chinese tech industry now? How would you compare it to Silicon Valley?

China is transitioning from a period of very fast growth into a period where the domestic competition is really fierce. At the same time, Chinese tech companies finally have a group of products that are able to compete internationally after about two decades of development. In the social media industry, TikTok is the first Chinese product that has gone mainstream elsewhere. And we are seeing the same thing in e-commerce, with Shein and Temu. In terms of electric vehicles, these companies are not in America for political reasons, but they are very successful in other countries. There are a lot of Chinese brands. So Chinese tech companies are pushed to go abroad because of the competition at home, and they have the ability to go abroad, but at the same time they’re running into political pushback in the US.

The arrival of DeepSeek took the US by storm, and caused a kind of panic. What do you think of that response?

China is able to come up with things that are new, that are useful, that are pushing the boundaries. Before DeepSeek, I also held the stereotype that the best students from China, the top scientists, the top researchers, tend to end up in the US. But actually the Chinese system still has a very strong talent pool. In that sense, it’s a reminder. But should America panic over that? Is that a political threat? I think people can debate that. 

Something I was hearing over and over with DeepSeek was that it produced a very comparable—maybe better—large language model to OpenAI’s on a lower budget and with dramatically fewer chips, calling into question the idea that you need more resources for more or better output. Do you have thoughts on that way of reading what’s happening here? 

I will agree that it challenges the concept of needing more chips to build better AI models. But it doesn’t mean that you don’t need those chips at all, or that companies like DeepSeek are already overtaking Silicon Valley companies in AI research. One reason DeepSeek was successful is that it started to acquire Nvidia chips very early on; it had its own stockpile of chips, and that’s what powered its early research. Even though later they said they figured out a way to train this model with fewer chips, it was built on previous research in Silicon Valley with a lot of chips. It’s a very natural thing in scientific research that someone else points in a different direction to explore, and then other companies start exploring that direction. In a way this is a community trying to work out different solutions.

Trump reversed the TikTok ban when he came into office, but then imposed tariffs on China. He has reversed AI regulations that Biden put in place. And Trump and Vance have used strongly protectionist rhetoric around keeping AI innovation within the US and characterizing the Chinese tech industry as an unwanted rival. Where do you anticipate things going from here? What are you going to be paying attention to in your coverage? 

We have already seen tariffs imposed over the past few years since Trump’s first presidency, and new chip bans. I am seeing creative ways to get around this, ranging from smuggling chips into China to simply passing the cost on to American consumers. Even after all the tariffs, we are seeing the rise of super-cheap e-commerce products flowing from China. I think it’s very hard to stop the business flow between these two countries, because there are so many goods being produced in China and China is eager to find a way to export them; Americans are eager to find a way to buy cheap stuff, and cheap stuff comes from China. Businesses on both sides will work together to navigate those regulations. In terms of AI, I feel it’s very hard to stop the flow of knowledge when you have smart people on both sides. They can see each other’s research outputs, even though you may be able to ban equipment or hardware. I think the DeepSeek case is a reminder that America is in some way pushing Chinese companies and people to get more creative, to find alternative paths that actually reduce their reliance on American hardware. 

Other notable stories:

• Yesterday morning, CNN’s Brian Stelter and Christian Edwards reported that several US-funded international broadcasters that are overseen by, but have some operational distance from, the US Agency for Global Media have continued to broadcast in recent days—despite a recent Trump administration edict seeking to gut them—on the grounds that the broadcasters’ leaders believe the edict to have been unlawful. Later in the day, one of those broadcasters, Radio Free Europe/Radio Liberty, went to court in a bid to block USAGM’s effort to terminate its federal grant, arguing that Congress already appropriated the funding. By contrast, Voice of America, which is more directly tied to the federal government, has fallen silent. For CJR, Liam Scott, who was just laid off from a job covering press freedom there, takes readers inside his final days at the broadcaster.

• Last week, CJR’s Sewell Chan and Betsy Morais reported that the prominent journalist Wesley Lowery had left positions at the Investigative Reporting Workshop and American University following at least three Title IX allegations against him, in which he was accused of improper behavior with colleagues and female students. (Lowery denied wrongdoing and said he left the posts voluntarily.) Now the Washington Post reports on the complaints and the account of a woman who felt that Lowery offered her a “sexual quid pro quo” over the future of a reporting contract. (Lowery disputed the account.) Elsewhere, The New Yorker cut ties with Jackson Arn, its art critic, after he was accused of making “inappropriate overtures” at the magazine’s anniversary party, per the Times.

• For Nieman Lab, Miranda Green tells the murky tale of a newspaper in North Dakota—with apparent ties to fossil fuel interests and a right-wing national media network—that has recently ramped up coverage about “the destructive nature and lawlessness” of protests against the Dakota Access Pipeline nearly a decade ago. The paper has circulated in a town where the company behind the project is suing Greenpeace over its role in the protests, a case that went to trial recently. Greenpeace has argued that the newspaper’s coverage formed part of a campaign to influence potential jurors and attempted to have the trial moved, but was unsuccessful.

• Il Foglio, a newspaper in Italy, claimed to have published the world’s first newspaper edition created entirely by artificial intelligence, part of a broader experiment aimed at exploring the impact of AI on journalists’ ways of working, The Guardian’s Angela Giuffrida reports. “The articles were structured, straightforward and clear, with no obvious grammatical errors,” Giuffrida writes. “However, none of the articles published in the news pages directly quote any human beings.” The edition did include AI-generated reader “letters” and editorial responses—one of which cast doubt on the reliability of AI.

• And Trump abruptly released a trove of documents about the Kennedy assassination to the public. Scholars said that it will take time to go through the files—not least because some of them are physically hard to read—but nothing explosive jumped out, with one academic telling the Times that the release had “convinced him that the information had been kept secret all these decades not because it included inflammatory information about the assassination, but rather to mask highly sensitive details about CIA intelligence-gathering.” A Times heading read: “Here’s what to know. (Oswald still did it.)”

Has America ever needed a media defender more than now? Help us by joining CJR today.