The advent of ChatGPT in 2022 sparked a global boom in generative artificial intelligence (AI). Its applications are rapidly expanding, and journalism is no exception. From AI-powered news anchors to assisted writing, editing, and transcription, image generation, and analysis of massive datasets, AI is now widely applied across every step of news production. Recently, newsroom-tailored AI consulting tools have emerged, aiming to promote “responsible AI use” among journalists.

In August 2023, numerous media organizations — including Agence-France Presse (AFP), the Associated Press (AP), European Pressphoto Agency (EPA), European Publishers Council (EPC), Gannett (the parent company of USA Today), and Getty Images — jointly issued an open letter calling for rules and responsible development principles for generative AI models. Soon after, the AP released its official AI guidelines, becoming the first major news organization to release AI use regulations for its newsroom.

In the Mandarin-language media sphere, Taiwan’s Central News Agency (CNA) and Public Television Service (PTS) published their own AI usage standards and guidelines in September 2023; The Reporter (報導者) , one of Taiwan’s top independent news outlets, followed with its own usage rules in July 2024. In China, the state-run China Media Group (CMG), the media conglomerate under the Central Propaganda Department, embedded political guidelines into its Interim Regulations on the Use of AI, introduced in March 2024, stressing that “adhering to correct guidance is always the foremost principle” and that “socialist core values” must be upheld regardless of technological development.

The AI-led transformation is impacting professional work cultures at some of the world’s oldest journalism outlets, including the Financial Times (FT), the British daily newspaper founded as a broadsheet in 1888. Like its English-language counterpart, the Chinese-language edition of the FT (FT中文网), broadly applies AI across its news production and distribution. It has launched its own chatbot and normalized AI integration into functions such as audio news readouts, newsletter production, image generation, and column translation. 

FT Chinese editor-in-chief Wang Feng (王丰), who also teaches journalism at Tsinghua University’s School of Journalism and Communication in Beijing, has witnessed the AI wave in journalism firsthand. Over the summer, he spoke with Tian Jian (田間), the China Media Project’s sister publication on Chinese-language journalism and media, to talk about how FT embraces AI technology while safeguarding journalistic professionalism — providing readers with more valuable services.

Tian Jian: FT Chinese makes extensive use of AI. From the reader’s perspective, the first thing that stands out is your dedicated chatbot, “ChatFTC,” whose name resembles ChatGPT. It acts as an AI financial news assistant for users. Could you share the process of the model’s development? What training data was used? And how have usage and reader feedback been so far?

Wang Feng: This wasn’t originally developed by FT Chinese. It’s a product created by the FT headquarters, where a technical team of over a hundred people built and pre-trained it. It’s based on the ChatGPT framework, but trained on FT’s news content. On that basis, we (FT Chinese) then retrained it again using our Chinese-language content.

Its special feature is that you can ask questions such as: what coverage has there been on this topic in FT’s reporting over the past year? What facts, what data? Give me a timeline, give me a long-term background introduction to this or that topic … In short, it allows an ordinary reader, beyond reading just one piece of news, to have a deeper experience, grounded in the FT news archive.

But I’m afraid it’s still difficult to completely avoid hallucinations. At least for us editors, the chatbot is still something of a black box. The tech team hands it to us after developing it, but we editors are not completely clear on how exactly they built and trained it, and to what extent it accesses the news database.

So we’ve explained to readers that this remains an experimental product. We hope it enhances their experience, but we cannot 100 percent guarantee accuracy. If someone needs reliable guidance for work or research, we recommend ultimately relying on FT-published articles as the standard.

Tian Jian: When did FT headquarters begin developing AI-related tools? Was there a delay for FT Chinese [compared to other newsrooms]?

Wang: FT now has a cooperation agreement with OpenAI. But in 2023 — before the OpenAI partnership — at the annual executive meeting at FT headquarters, all executives were required to immediately register for an Anthropic account and a Claude AI account. The idea was to press us to see firsthand how much AI had already advanced.

As FT executives, that was the first time we realized large language models had made such dramatic progress in text. Around the same time, the London headquarters began negotiating with AI companies, including Anthropic and OpenAI, and in 2023 and 2024, they began in-depth development. FT Chinese progressed in parallel — once HQ has built something decent, we can adopt it without delay.

Tian Jian: How are FT users currently using the AI financial news assistant “ChatFTC”?

Wang: Of course, we can’t see exactly what readers are doing with it. But the tech team can see how many people are using it, how much traffic it consumes, and roughly how frequently each reader engages. They regularly pass this data back to the editorial team.

But at present it’s more of an interesting novelty; the usage rate is not very high. One major issue is that ChatGPT is inaccessible in mainland China. Even for myself, though I’m based in Hong Kong, I have to use a VPN to access it. Readers in the mainland need to scale the firewall in order to use this tool, which is very difficult. It’s also limited to subscribers. So there are multiple levels of restriction, meaning perhaps only the most tech-savvy [mainland] readers who know how to successfully bypass restrictions end up using it.

Tian Jian: How is the readership for FT Chinese distributed?

Wang: Our primary market is mainland China, which makes up the overwhelming majority: 85 percent of readers are in the mainland. Hong Kong accounts for 5-8 percent, followed by Taiwan and Singapore, and then overseas Chinese readers in North America, Europe, the UK, Australia, and Canada. FT Chinese was originally created as a Chinese-language product geared toward mainland China. When we started 20 years ago, the idea was to translate FT’s English content into Chinese for those in the mainland who had very limited access to international information.

Tian Jian: From a reader-centric standpoint, when it comes to products like chatbots, has the team considered integrating DeepSeek? What experience does your team have with Chinese-developed AI tools?

Wang: This touches on the geopolitical regulatory problems with large AI models. Since our parent company (FT) is a British company, and has already signed a cooperation agreement with OpenAI, the company recognizes it as sufficiently safe. The current company policy is to mainly use OpenAI’s products for AI applications. On top of that, because cross-border data flow issues under the EU’s GDPR (General Data Protection Regulation) affect China, the UK, and Europe, even if FT Chinese wanted to collaborate with DeepSeek, it’s unlikely that headquarters would approve.

Our staff currently doesn’t have very strict limits on using Chinese tools. But when transcribing interviews, if sensitive information is involved, can we upload it to iFlytek’s servers? If colleagues are concerned, we discuss it together. My advice is: if it’s sensitive, don’t use iFlytek. Transcribe it manually. Luckily, most of our content doesn’t reach that level of sensitivity.

My colleagues and I use DeepSeek, Tencent Yuanbao, and Doubao quite a bit. We find DeepSeek’s most significant flaw is hallucination: it often presents fabricated information in a serious manner. This happens quite often, more often than with Perplexity or ChatGPT. So I personally rarely use DeepSeek for interviews or content-related work. That said, its proficiency in Chinese is much smoother, so sometimes I’ll use it to rewrite headlines or write summaries — but I always re-read to ensure accuracy.

FT recently began conducting surveys to better understand which AI tools everyone is using. I suspect the next step is to formulate a strategy to prevent possible data or privacy issues. We are making up the rules as we go. Internal management policies are continuously being updated.

Tian Jian: Does FT headquarters currently have a set of guidelines or rules for AI use?

Wang: They have some basic codes of conduct, and employees are required to undergo regular annual training and internal examinations. But enforcement is difficult — no company can really monitor every single task each employee does.

In daily work, the way these principles are applied comes down to individual judgment. If you’re unsure about something, you talk to your manager for guidance. If the supervisor can’t decide, it escalates to consultation with a legal counsel or an in-house lawyer.

I often discuss these issues with my colleagues and share my own AI usage experiences with my team. The most basic principle is: even if AI generates something, you are still 100 percent responsible for its accuracy. You must verify everything it tells you.

Right now, most people are still exploring how to use AI. Few are thinking far enough about how to ensure AI outputs comply with all professional, ethical, and legal standards. Even just ensuring 100 percent accuracy is already quite difficult. We are still in the very early stages.

Tian Jian: Aside from ChatFTC, is FT also using AI in podcast production?

Wang: Many of these are still experiments and have not yet been officially launched. For instance, people found it interesting that NotebookLM could create podcasts, but after a few days of using it, they realized it’s not possible to control how issues are discussed, so the final results aren’t good enough to present to readers as a news product. It’s more useful for internal brainstorming. Many AI tools are like that: fun to try, but we haven’t figured out exactly how to apply them yet.

Where FT Chinese has applied AI in its news products more systematically is in voice summaries. For example, turning a 2,000-word article into a 200-word summary, then using an AI voice to read it aloud, and sending it out daily to readers via WeChat. This is useful for drivers or commuters who can listen to articles with their earphones.

I think Chinese readers are very open to new things. They’ve responded very well to AI voice summaries. But I once tested an AI podcast on my personal WeChat account. As soon as readers were told it was just two AI voices talking to each other and was made by AI, their interest declined. So the key is still the added value of humans.

Tian Jian: Reporters and editors are responsible for their output. Has FT Chinese ever encountered problems with AI use?

Wang: We haven’t had any scandals or major factual errors. Our internal practice is that original editorial content passes through at least two pairs of eyes. Compared with traditional media, where three or four people review drafts, this may leave more room for error, but so far we haven’t encountered obvious problems.

Currently, FT Chinese uses AI tools mainly for peripheral tasks, especially on the content distribution side. When producing original content, we still abide by tradition: reporters conduct face-to-face or video interviews, record everything, then turn it into text, video, or audio — a very traditional content creation process.

Where does AI fit into this process? For instance, when interviewing an expert in a field we know little about, how can we come up with good questions? This is an area where we will use AI more often. The interview itself is still conducted face-to-face, and what the interviewee says must still be noted and written down by us.

If AI helps with transcription, we still need to verify the content afterward. This process already filters out most inaccuracies or hallucinations that AI might produce. So AI plays a supporting role.

Once we’ve produced the text or video, we can then use AI to further process it — turn it into a newsletter or a podcast. In other words, content is produced through a traditional, verifiable journalistic process, and then it’s further produced or distributed with the help of AI.

Tian Jian:  Does this show that journalism as a profession still has value? How do you view today’s flood of AIGC (AI-generated content) online?

Wang: Articles written by AI absolutely cannot be published directly. Even after multiple rounds of editing, we’re not necessarily confident that AI-generated drafts can have accurate journalistic judgment.

Our priority is to ensure as much as possible that in the core workflow of journalism – especially when it comes to originality, accuracy, objectivity, and newsworthiness – there is always human judgment. If we can ensure that, then the resulting content won’t differ much from what traditional processes produce. AI is only used for secondary or tertiary processing, improving efficiency in distribution. This way, we’re less worried about possible inaccuracies.

On YouTube and TikTok, you already see massive amounts of AIGC. Much of it is entertainment or leisure content, with low information value, not much different from the social media “clickbait” era before AIGC. Aside from wasting huge amounts of people’s time and causing “brainrot,” it’s not a major direct harm.

But content that lies between entertainment and news and information, large volumes of AI-generated content that is unverified, contains false or incomplete information, and is spread widely — that can cause serious harm. Political, commercial, or criminal actors can weaponize it for misinformation or disinformation campaigns. The risks are many times greater than in the social media era and are deserving of serious vigilance. Some countries, including China, have already created legislation requiring social and information platforms to label AI-generated or AI-assisted content. I think that helps mitigate the harms that come with AIGC.

Tian Jian: What AI skills should journalists develop? Which AI tools do you most recommend?

Wang: Personally, my favorite is Perplexity. It’s very powerful, and even the free version can already examine a lot of content. For handling large volumes of data, NotebookLM is quite useful, since it ensures no fabrications. At the moment, it’s capable of handling cases like dozens of presentation slides or a few hundred thousand-word manuscripts or academic papers.

Another is Google Pinpoint. I’m still learning it and haven’t found the right opportunity to apply it. My understanding is that it’s well-suited for combing out datasets on the scale of something like the Panama Papers, when you need to sort through and search for leads. But we haven’t had the chance to test this tool with such large datasets yet.

There’s also Manus, currently the only AI tool I personally pay for. It can handle very complex tasks. Recently, I wrote a journalism textbook based on my past two years of teaching part-time at Tsinghua University (in Beijing) and the University of Hong Kong. But I didn’t write it manually: I uploaded 600–700 pages of PowerPoint slides to the AI and used it to write.

I tried many tools, including ChatGPT, DeepSeek, and Manus. After months of trial and error, I worked out a process. First, I had Manus create a clear chapter structure according to the slides, essentially creating the framework of the book. Then I used ChatGPT’s Deep Research function to fact-check, supplement content, and connect the logic within the text. After researching this process, I could write the book much more efficiently.

I give my students two basic principles. First, I encourage them, even require them, to use AI. Because if they don’t, by the time they graduate, everyone else will be using it, and they’ll lack competitiveness. Second, I tell them that they must take responsibility for all content, whether generated by AI or not. If AI produces hallucinations, bias, or inaccuracies, you must be able to detect and correct them.

And how do you detect and correct them? This is why we still need traditional journalism education. You still need to learn today about what journalism was 20 years ago, so you can know what AI gets wrong and what it gets right. Without that, you don’t have the ability to make judgments.

Tian Jian: For students using AI, do you set limits or quotas?

Wang: I don’t set any limits. Using a tool to determine whether 70 percent of an article was written by AI is a foolish task. So-called “AI-detection tools” are unreliable and always lag behind large language models.

Now, many students worry that, even if they wrote their papers themselves, tools may still say it’s “40 percent AI.” So they spend more time worrying about being mistaken for using AI to cheat. That’s completely counterintuitive. So I see no point in restrictions.

My view is: no matter what tools are used, if it’s well-written, that’s great. As long as you can ensure it’s accurate, I’ll give it a good grade.

Tian Jian: What advantages does the growing use of AI and social media among the next generation of journalists and journalism students bring to the profession?

Wang: At Tsinghua, I teach traditional journalism courses, English news writing, and business news writing. But every day I ask my students: where in the classroom could you use AI to be faster? Or, if you’re comparing financial reports from the same company across two years, which tool helps you be more accurate? I use AI tools in this way daily to reinterpret traditional journalism workflows.

AI is particularly helpful for journalists and students like us whose native language isn’t English; it’s particularly useful in helping us master writing style and correct grammar. Over the past two years, I’ve seen students’ writing improve a lot. The content they hand in is quite good. But I can’t be sure how much of that they learned themselves and how many shortcuts were made possible by AI. 

I can also be sure that I must teach them how traditional journalism is done, what’s right and what’s wrong, what’s good. Even if you didn’t write it yourself, at the very least, you need the ability to make a judgment. Using AI to help with processing is fine — ideally, you would add your own insight and perspective during the process. Ultimately, the content must be accurate, meet traditional journalistic standards, be good, fresh, and sustainable. If you can do that, it’s enough.

Tian Jian: I’m curious, how do you think Western and Chinese-language media differ in their attitudes toward AI?

Wang: At least from what I’ve seen, journalists and readers in mainland China tend to be more open-minded. Of course, we must ensure everyone recognizes the risks, including the ethical, legal, privacy, and data security problems. But overall, compared with the West, China seems more accepting.

Within FT, I’m personally more cautious. Many of our senior journalists and editors are not so quick to embrace these tools. It’s still a challenge for the company to persuade them to adopt AI more fully.

Tian Jian: You seem curious, experimental, and enthusiastic about new technologies yourself. How about the rest of your FT Chinese team — what’s their general attitude toward AI?

Wang: I am indeed especially interested. I’ve been a journalist for 26 years across wire services, newspapers, and magazines, but for about 20 of those years, I’ve been an editor, mostly working on digital platforms. So I’ve always been something of an internet native, always relaunching websites, building new functions for the website, and adding social media features. Maybe that’s why I’m more open to AI.

To some extent, I influence my students more than my colleagues. For colleagues, all I can say is: “Here’s a process I figured out that will make your work simpler.” For example, when COVID started five years ago, the company carried out major layoffs. Suddenly, our staff had been reduced by a lot, but the workload was just as much as before, and everyone suddenly faced enormous pressure.

During that process, I helped them think through ways AI could speed up transcription. I drafted an initial workflow, then we discussed it together so they could see the practical benefits. Otherwise, they would be working 15-hour days. This way, they were more willing to buy in.

I never forced anyone, nor did I set KPIs or deadlines for AI adoption. I just showed them that sticking with the old ways was exhausting, while AI could boost efficiency; is this not a good thing?

Tian Jian: When reporting on China-related topics, how does FT Chinese strike a balance that maintains professionalism and depth, while also ensuring smooth publication? Have you encountered subjects that require extra caution?

Wang: FT Chinese’s mission is to serve China’s business and professional readers with economic, financial, and technology information. This positioning helps us resolve many potential legal, policy, and regulatory issues we may encounter in the Chinese market. At the same time, we emphasize showing the Chinese perspective, inviting many mainland and Greater China experts, scholars, and professionals to write original commentary and analysis. This helps us balance Chinese and Western discourse and provide readers with professional, neutral, and balanced in-depth information.

Tian Jian: On June 28, US financial weekly Barron’s partnered with Chinese financial-tech media TMTPost (鈦媒體 ) to launch Barron’s Chinese (巴倫中文網). How do you view Barron’s strategy to launch Chinese-language content at this moment?

Wang: Barron’s has a competitive relationship with FT, so naturally, Barron’s Chinese will also compete with FT Chinese in both the content and business aspects. Given that many international media platforms’ Mandarin-language sites have shut down or withdrawn from the mainland in recent years, we view other international outlets opening Chinese platforms to serve mainland readers as a positive development.

This interview was translated and edited by Jordyn Haime, with assistance from Claude AI.

In the late 1990s, the media landscape in China was overtaken by a wave of commercialization and marketization as the Chinese government sought new ways to support otherwise expensive newspaper and broadcasting operations — and to encourage a “media industry” (an entirely new concept at the time) that was more suited to the country’s rapidly developing economy. One after another, media companies launched market-oriented reforms. And while media groups, most linked to provincial and city governance structures, pursued economic benefits, many working within these emerging outlets began embracing something less expected: journalistic professionalism. 

The new generation of media outlets quickly distinguished themselves from their previous roles as propaganda outlets, serving growing, and increasingly affluent, audiences. Perhaps the most representative of these changes was the Nanfang Media Group (南方報業傳媒集團), which was known at the time for its suite of professional media outlets, including Southern Weekly (南方周末), was once praised by The New York Times as “China’s most influential liberal newspaper.” During this period, many liberal intellectuals used these media platforms to disseminate new ideas, and to speak more critically on social and political issues. 

Veteran journalist Li Sipan (李思磐) joined the Nanfang Media Group in 2002, working there in various roles for a decade. But even in the space afforded by this relatively free and open environment, she keenly felt frustration at the systematic neglect of women’s rights. Later, inspired by Sun Yat-sen University Professor Ai Xiaoming (艾晓明), a women’s rights activist and filmmaker, Li Sipan and a group of female media professionals established the feminist platform Women Awakening Network (新媒體女性). They conducted advocacy on women’s issues, organized training workshops for female journalists, and hosted exhibitions and lectures — all while actively producing critical reporting on gender-related topics.

The period of relative openness proved to be short-lived. Following a cascade of events — including the 2013 controversy over the censoring of the New Year special issue at Southern Weekly (南方周末), and the 2015 “Feminist Five” incident — China’s market-oriented media experienced a rapidly shrinking public discourse space. Feminist voices came under severe suppression, and the Women Awakening Network also faced mounting pressure. Li Sipan was forced to leave the NGO in 2018 to take a university teaching job. On May 21, 2021, the Women Awakening Network Weibo account suspended regular updates. Its final post was a repost about the Zhu Jun (朱軍) sexual harassment case.

Li Sipan sat down with Tian Jian (田間), the China Media Project’s Chinese-language outlet on journalism and media, and CMP researcher Dalia Parete to discuss the absence of feminist consciousness in China’s liberal media. Speaking as both a feminist activist and an investigative journalist, she offered her observations and insights on China’s feminist movement and women’s journalism since the 2000s.

Tian Jian/CMP: Chinese media institutions were historically male-dominated. Did you face any challenges as a woman?

Li Sipan:  I started working in the investigative department of Southern Metropolis Daily (南方都市報) in 2007. At that time, there were only two female reporters in the department — one in Guangzhou and another at a different bureau. The female reporter in Guangzhou was happy to have another woman join, noting that the constant presence of a large group of men in a smoke-filled room had made her feel very uncomfortable.

We didn’t have to work fixed office hours, and back then, every Monday, those of us who weren’t traveling for work would have dinner together. We were a really tight-knit group — everyone was good friends, and our work and personal lives overlapped quite a bit. But as a woman, I still felt a bit isolated. When we’d eat together, it was a male-dominated scene where the guys would drink heavily, competing to see who could drink more until they were making fools of themselves. Of course, the younger generation probably doesn’t do this anymore. But the way the men talked to each other often left us feeling awkward and quiet. They would casually toss around words like “beautiful” or “sexy” and other random, totally inappropriate words to refer to us.

This divide was also quite evident professionally. During the era of market-oriented journalism, social affairs and legal reporting were important beats that could easily enhance a journalist’s reputation. Since this type of reporting often involved cases of wrongful conviction and judicial injustice, it aligned with the traditional Chinese ideal that intellectuals shouldering moral responsibility and upholding justice, making it easier to capture readers’ attention and achieve notable results.

This type of reporting was typically easier for men to excel in. Public opinion supervision at that time was primarily conducted through cross-regional reporting, which required dealing with local officials. There was a common understanding that male reporters could drink with these officials until they became like sworn brothers — after some chest-thumping and shoulder-patting, they might receive news tips or key documents. I would also engage with officials, but honestly, many of these official drinking sessions were quite crude, with rampant sexual harassment, dirty jokes, and inappropriate physical contact. This made things really difficult for women, so I preferred to accept less-than-perfect results rather than drink with male officials.

The kinds of stories we covered back then were completely different from today’s. A large proportion of coverage focused on rural issues, whereas now approximately 90 percent of topics center on the urban middle class.

What’s particularly interesting is that many of the male reporters in the investigative department back then were quite legendary figures. Some had previously sold fruit for a living, others had never formally attended university, but they excelled at breaking through barriers and “getting the scoop.” However, the female reporters had somewhat different backgrounds—the women around us typically had stronger educational credentials. Most of the men held bachelor’s degrees, while quite a few of the women had master’s degrees. I originally joined the investigative department at Southern Metropolis Daily because the department head wanted someone who could cover intellectual affairs, civil society issues, and Greater China stories. They liked that I knew Hong Kong, Macau, Taiwan, and Singapore, so I even spent some time reporting on Taiwan. As the Global Media Monitoring Project has observed, although I also covered politics as a female reporter, my assignments were still relatively “soft” — culture and society oriented — compared to the hard-hitting stories assigned to male reporters.

TJ/CMP: Why did you become so focused on women’s issues?

Li: It’s a long story. Feminism has appealed to me since middle school, probably. When I was in university and they held the World Conference on Women in Beijing in 1995, I wrote to the organizing committee asking for materials. My writing and studies were related to women’s issues, too. But what pushed me to take action was dealing with this male-dominated newsroom culture.

Nanfang Media Group was considered the most progressive media organization at that time. What impressed me about them was this: when I was in Shanghai, local newspapers would refer to migrant workers as “blindly flowing outsiders” (外來盲流), which I found to be an extremely discriminatory term. But when I arrived in Guangzhou and began reading Southern Metropolis Daily, I noticed they didn’t treat migrant workers as a special or marginalized group. Their reporting would provide detailed portraits of individual migrant workers, including where they came from, what kind of work they did, which factory they worked at, and so on.

Later, I realized that even in Guangzhou’s press industry — which seemed to be the freest in China — there were many gender-related aspects that made me feel uncomfortable. For example, in 2005, one district implemented gender education where, during exercise breaks, girls were required to dance while boys practiced martial arts. Several reputable newspapers in Guangzhou reported this as a progressive development.

There was also the Peking University minority language program admissions controversy, where parents protested because female students needed admission scores dozens of points higher than male students. At the time, Southern Metropolis Daily‘s editorial department was extremely progressive — constantly discussing political reform — and yet they published a commentary arguing this wasn’t gender discrimination. Also in 2005, when China was amending the Law on the Protection of Women’s Rights and Interests (婦女權利保障法), women’s organizations worked tremendously hard to include anti-sexual harassment clauses. But Southern Metropolis Daily basically said it was ridiculous —“wu li tou”— like something out of a Stephen Chow (周星馳) martial arts comedy.

I sincerely believed in the Nanfang Media Group philosophy. It provided journalists with tremendous freedom. I had a lot of personal growth there, and I formed some of the most important friendships of my life. But regarding gender issues, it was problematic in so many ways. Southern Metropolis Daily had several female editorial board members, but the journalistic culture remained very male-centered. Even the female leaders lacked sensitivity to gender issues. Like leaders in many mainstream institutions, while these women may have demonstrated more empathetic leadership styles compared to men, professionally they felt compelled to perform after the pattern of their male counterparts  —perhaps even consciously or unconsciously distancing themselves from their female identities. In essence, everyone considered women’s rights unimportant, viewing gender equality as communist overreach — a failed ideological agenda.

Actually, I preferred being a reporter, but I started writing opinion pieces because of the Peking University minor languages program incident. I couldn’t sit still without writing about it, and that later became my standard response whenever I wrote commentaries. I wrote the piece, but Southern Metropolis Daily wouldn’t publish it, so I took it to Shanghai’s Oriental Morning Post (東方早報) instead. The editor there said I wrote well, but he also said [giving his sense of why it couldn’t be published]: “What you have there isn’t liberalism — it’s anti-communist.” 

TJ/CMP: How did Women Awakening Network come about?

Li: The turning point was getting to know Professor Ai Xiaoming. Professor Ai was working tirelessly to transform the media reporting culture at the time. She once organized a seminar specifically about Southern Weekly’s misogynistic advertisements and invited the then editor-in-chief Xiangxi (向熹) to participate. But in reality, no one in the media industry would actually endorse that kind of approach.

Later, Professor Ai collaborated with the British Council to bring in BBC experts for training on media and gender. The establishment of Women Awakening Network (新媒體女性) was a direct result of this training initiative. I wasn’t the sole founder. We were a group consisting of 12 media professionals from Guangzhou who had participated in the training.

Actually, that training had some issues. Because we needed interpreters for everything, the two-day workshop couldn’t cover much ground. Additionally, the BBC has these ethical guidelines, the kind that are necessary from both a legal standpoint and for social responsibility in places where there is actually freedom of speech.

But matters of journalistic ethics are hard to push in China. Chinese journalists really hate the whole journalism ethics thing, mainly because China is a place where it’s tough to get public information out there in the first place. Journalists face all kinds of obstacles and risks. If you can manage to get a story published — that’s already something. 

So when we [at Women Awakening Network] advocated for gender mainstreaming, we didn’t position journalism ethics as our core training objective. We used the term “journalistic professionalism” (新聞專業主義) instead, focusing on providing journalists with more gender experts and activists as news sources — helping them develop their stories rather than telling them not to pursue certain types of coverage. We trained them on applying public service news values in their reporting, revealing truth, and promoting reform, based on the existing operating logic of market-oriented media.

Chinese women’s rights groups were unfamiliar with commercial media due to their operational methods. I noticed that the press avoided covering women’s issues because the two sides rarely interacted — in contrast to human rights lawyers and environmental groups at that time. So we wrote a few contact lists for women’s research and activism groups. We helped researchers, activists, and journalists connect. We provided organizations with lists of journalists they could contact, and journalists with lists of experts and groups they could call.

TJ/CMP: What was unique about Women Awakening Network‘s feminist activities?

Li: Before 2003, Guangzhou had no feminist organizations. In the wake of the World Conference on Women, a lot of feminist organizations were established in Beijing, as well as in Zhengzhou and Xi’an, building on the groundwork laid by previous generations of feminists in those places. So when international funding became available, they naturally prioritized investing in and developing these existing networks first.

At that time, feminists from northern China would often describe Guangzhou’s feminist activities as “very spirited.” Why did they say this? Because the leaders of women’s rights organizations in Beijing tended to come from official media outlets, or from the Women’s Federation (婦聯), or to be scholars from government think tanks. They were all people within the system. They not only needed to maintain good relationships with the government — they also had to consider the government’s operational logic, trying not to cause trouble. That is to say, rather than exposing social problems, they had to be helpful by promoting women’s rights concepts through publicizing the political achievements of local governments that were willing to support feminist projects.

But Guangzhou was different. We served as Women’s Federation experts for a period under Wang Yang’s (汪洋) governance of Guangdong, but our work wasn’t directly connected to the institutional Women’s Federation operations. Guangzhou is a media hub, so we used the media as our point of leverage. At that time, Guangzhou hosted all sorts of activities on remarkably progressive topics on a weekly basis. Public intellectuals were quite bold in the things they said, things that would be nearly impossible to hear publicly in other Chinese cities.

But there were hardly any female speakers back then. There were no talks about women’s issues — and definitely no feminist lectures. So we started organizing talks, exhibitions, seminars, and other events. We would invite journalists from the media to come. Commercial media journalists generally tend to resist being lectured to, or having ideas pushed on them. Instead, you need to let feminism exist in their city, and make feminist voices a real part of the conversation.

By the time the young feminist activists emerged in 2012, Guangzhou media already had a group of journalists who were interested in feminist issues and had a basic understanding of women’s rights organizations. This was the result of our decade of work. Therefore, Guangzhou media was the most supportive of young feminist activism. Of course, at that time feminism wasn’t the only movement flourishing. Guangzhou’s civil society was vibrant on many fronts. Young people were engaged in a range of causes, from environmental protection to budget transparency, cultural preservation, and so on. Our work in women’s rights also influenced a certain cohort of young people.

TJ/CMP: Women Awakening Network also worked with online platforms like NetEase (網易) and Phoenix.com (鳳凰網), putting out some of the in-depth stories. Why did you decide to partner with new media?

Li: In 2014, when the Wu Chunming (吳春明) sexual harassment case happened at Xiamen University, we launched an eight months-long campaign around it. We supported the victims, provided them with legal advocacy, and ensured their voices were heard in the media. We published a series of investigative pieces and detailed commentaries about the case. We collaborated with Chinese university teachers and scholars from around the world to translate anti-sexual harassment policies from various countries and universities. We also drafted policy recommendations for the Ministry of Education and Xiamen University, and secured the signatures of scholars around the world on joint letters to the Ministry of Education. We also put together sexual harassment prevention handbooks for college freshmen, organized film screenings about sexual harassment in academia for lawyers, social workers, and people from all walks of life — that sort of thing.

By that time, the advertising market for newspapers was already collapsing. After the 2013 Southern Weekend incident, investigative reporting departments were shuttered and many veteran journalists were leaving the profession. As experienced professionals left traditional outlets, many simply lost their voice. Most of the resources had shifted to online platforms. As competition between news apps heated up, online portal sites like Sohu, NetEase, Tencent, and Phoenix all started doing original news content for their apps. And they were poaching a lot of talent from traditional media, offering much higher salaries.

Internet companies had abundant resources, and they emphasized efficiency. They wanted to leverage their platforms to mobilize more resources. Therefore, they were very open and flexible. For example, when our grassroots advocacy group for legislation against domestic violence required media support, all the major portals provided us with resources to raise our public exposure — things like live streaming and expert interviews. During the Xiamen University incident, the departure of many veteran journalists left traditional media without the talent they needed to cover the story properly, leading to lots of mistakes. Internet companies were more flexible and could accommodate outside sources, such as people from NGOs, to collaborate on news content production. An outlet like Southern Weekly probably couldn’t publish an article by a women’s rights organization.

We worked with NetEase’s Plum News (真話頻道) to put out reports about the Xiamen University incident, which helped get the word out.

But relative to internet outlets, traditional news media still have a stronger watchdog role. This is because they are not purely commercial. In a way, they’re part of the propaganda system. Eventually, the Ministry of Education put out “Opinions on Establishing and Improving Long-term Mechanisms for University Teacher Ethics Construction.” This was the first Ministry of Education document that banned sexual harassment. At that point, Xiamen University was still dragging its feet on taking action against Wu Chunming (吳春明), the perpetrator in that case. So we arranged for two of the victims to appear on “Oriental Live Studio” (東方直播室), a program on Shanghai’s Dragon TV, where they could share their stories. The day after the program aired, Xiamen University finally announced they were stripping Wu Chunming of his Party membership and his teaching credentials. But the decline of traditional channels was about more than just money and changing technologies. Soon after that segment aired, “Oriental Live Studio” was shut down [for political reasons].

Internet platforms and traditional media each had their distinct roles. However, from 2015 to 2017, the traditional media space shrank rapidly, while social media also came under state control. For instance, in 2016, we organized a rather interesting campaign to resist March 7 — a feminist critique of the official “Girls Day” and its often sexist rituals. The “anti-March 7th” (反三七) opposed the official depoliticization of International Women’s Day, its detachment from women’s rights, and the support for or tolerance of “Girls’ Day” (女生节) — a campus tradition that reinforces gender stereotypes, discrimination, objectification, and the sexualization of women. In contrast, the call to “celebrate March 8th” (过三八) emphasized women’s civic identity and demanded gender equality, especially within the university context.

Still, by 2017, we could no longer do it because I personally experienced quite severe online harassment. Of course, today this type of [online harassment] is directed at activists and journalists across the board. Later on, due to the passage of China’s NGO law, NGO work became very high-risk. I was “advised” by [government agencies] and gently urged to exit the Women Awakening Network organization. That turned me toward teaching.

TJ/CMP: Given the current situation with such strict government control, how can feminist issues and movements be promoted and spread?

Li:I personally focus on women’s rights and journalism, and both areas have become extremely challenging since the pandemic.

At the beginning of the pandemic, many journalists were still actively doing reporting. Some were citizen journalists who weren’t affiliated with any news organization, but when they went to Wuhan, many of them were detained and arrested. Back then, we still felt journalists could still manage to do meaningful work through non-fiction platforms and other alternative institutional media (机构媒体). But today, journalists from established media outlets are facing the same fate as those citizen journalists back at that time. Everyone has become extremely cautious, and it has become very common for journalists to face both state violence and online violence.

And don’t even get me started on women’s rights. The feminist movement has actually been systematically suppressed step by step, from the 2015 “Feminist Five” incident right up to now. At this point, it’s basically impossible to have any registered women’s rights organizations or advocacy groups. New Media Women held out for a long time, but they were also forced to shut down between 2020 and 2022.

The [authorities] invest a lot of energy in controlling the spread of feminism. They will say to individuals and organizations: “Look, there are some things you can do, but don’t talk about them online.”

They control the dissemination of public and feminist-related information. What is meant here by “public”? It’s often connected to government authority, but it may also relate to the breadth of information transmission (傳播範圍擴大). Suppose you discuss domestic violence, bride prices, or marriage and family issues that generate significant controversy. In that case, they might say you’re inciting gender antagonism. This is what ordinary bloggers encounter. But you can still engage in these discussions. However, anything related to ideology or state institutions is strictly guarded against. For instance, the Liu Qiangdong (劉強東) sexual harassment case can still be discussed, but the Zhu Jun (朱軍) case [involving a prominent state media figure] cannot. 

I once made a social media post discussing how many women had been killed in the war in Ukraine, and then the police contacted me. Initially, I thought it was because of the Ukraine war, but it wasn’t. Their reasoning was that it involved women, because that article had garnered hundreds of thousands of views. Given how vigilant they are about the spread of feminism-related information, launching feminist advocacy campaigns as previously defined has become extremely difficult. For example, during the chained woman incident, some female netizens took personal action to visit that village. 

But the undeniable fact is that today, after years of advocacy and awareness-raising, feminism has become common knowledge. It no longer requires organized feminist groups to drive its influence. It’s more like individual water droplets converging into an ocean, with many different forces now working to spread feminist ideas. However, we no longer have the conditions for the kind of organized activism we had before, nor are we likely to see mainstream media rallying together to cover and support a centralized feminist movement.

Sina Weibo came under much greater scrutiny in 2012, and the citizen activism accelerated by Weibo, especially liberal citizen activism, became the primary target. Of course, the marginalization of public intellectuals that came with the crackdown brought the feminist movement two or three years of relative visibility on social media. But while feminist discourse still appears quite “mainstream” on social media today, the reality is that the current feminist discourse is a result of the feminist movement being systematically eliminated from the communication sphere. Therefore, I believe we can no longer rely on a digital space that is heavily subject to censorship and algorithmic control.

Feminism serves as a very important binding force. For building alternative cultural spaces and non-mainstream communities, the critical perspective of feminist theory is extremely valuable.

If we view only  these algorithm-influenced platforms, the feminist thought on them is extremely uniform. To use an academic term, it could be said to have very strong neoliberal characteristics. Due to the absence of face-to-face connections, feminism has become a popular catchword used by women of relatively higher social status to legitimize their self-worth and advantaged positions. It is no longer based on public participation and accountability as starting points. 

Algorithms and commercial interests might not eliminate your connection to the public. For example, when we organized our “Resist March 7, Celebrate March 8” campaign, the platform later gave me a “Weibo Big V Award.” But censorship breaks the connection. It systematically dismantles trust between people and undermines our capacity for collective action. So if the feminist movement wants to accomplish anything today, we need to step away from social media platforms and connect with real people again. Like in the early days, we should organize small in-person gatherings that foster genuine face-to-face human connections.

TJ/CMP: From your point of view, how does China’s current media environment differ from that of before? 

Li: It’s clear now that there are more female journalists doing in-depth reporting, around the same age as my students. But they haven’t experienced the hardcore news-reporting environment of the “golden age of journalism,” so I also feel that the way young journalists approach reporting has become quite different.

This has to do with the influence of political censorship, business models, and other factors. You’ll also notice that rural topics have become less common. Many young journalists grew up reading in-depth reporting that was primarily in a creative nonfiction style, and the in-depth pieces they write also lean toward that creative nonfiction approach. Of course, creative nonfiction is itself a way of pushing back against the censorship environment. But for example, when someone asks me to look over a draft, I’ll say that back when we worked at newspapers, this article would have been 5,000 words at most  —how did you end up writing 12,000? They tend to focus on literary writing and include a lot of details that we used to think were unimportant in journalism. But sometimes, parts of the fundamental news elements — the 5Ws and 1H — are missing, creating this kind of suspended state where time and place feel disconnected.

As for what’s possible, even media outlets with editorial rights (採編權) — [meaning they are authorized to conduct reporting] — in many cases no longer do investigative reporting or engage in watchdog journalism. Some long-form nonfiction or feature-oriented media outlets are perhaps still trying to find ways to carry out such work, but they cannot provide journalists with the necessary conditions to do it properly, including, in some cases, even the legitimate press credentials required to report.

TJ/CMP: In your article News Media and the Feminist Movement in China: A Brief History, you wrote that female journalists have their own community, noting how they supported each other during the Zhuhai hit-and-run incident of 2024, for example. Can this compensate to some extent for the industry-wide issues you mentioned?

Li: I think journalists all grow up with the company and competition of their peers. Actually, there’s more cooperation than competition, so it’s very good for female journalists to have a network. In the past, there might have been more media reports, and everyone’s cooperation was like scattered flowers — different provinces all had media reporting and publishing reports. But now only a few individual media outlets can publish much of anything — even if everyone’s level of cooperation remains the same.

TJ/CMP: After 2017, you went into teaching. What advice would you give to students or young Chinese people who want to become journalists?

Li: I previously taught domestically at Shantou University in Guangdong. Shantou University’s Cheung Kong School of Journalism and Communication was greatly influenced by its founding dean, Ying Chan [NOTE: Chan was also the founder of the China Media Project]. So it emphasized journalism practice and valued teachers’ newspaper industry backgrounds. But later it became much like other journalism schools in China, where teachers must have doctoral degrees [over practical experience]. I felt at the time that this was problematic.

On the other hand, some of my students were intimidated or interrogated by the police while still serving as interns. For all sorts of reasons, journalism has become extremely high-risk. The risks they face are completely different from what we faced back then. At that time, we had institutional protection. Now, even with institutional protection, it’s not really possible to do real journalism. 

But I still believe in journalism. The skills you develop as a journalist—learning to explore and understand the world on your own terms—will serve you no matter what path you take later. And for young people willing to pursue this work even in today’s harsh environment, driven by idealism and a desire for justice, that persistence will pay off. They will find ways to make a difference.