“A small plane hit CITIC Tower??? How could something like this happen? What happened to the air defense capabilities in our core restricted zone???“ one Weibo commenter wrote on Friday night.

After videos and images of the incident spread, online discussions were swiftly and thoroughly censored. That same day, while media outlets around the world began reporting on the crash, it did not appear among the top 50 trending topics on Chinese platforms—in fact, it wasn’t reported anywhere at all. As related keywords and hashtags were taken offline, it quickly became clear that this was the “6.26 Incident That Shall Not Be Mentioned.”

Even the next day, when Chaoyang authorities issued a brief official notice, republished by state media, the name of the tower was never mentioned, just that “a single-engine, two-seat light sport aircraft collided with a high-rise building during flight,” and that an investigation was underway. The one person on board, the pilot, died in the crash. Thirteen others were injured.

Later, flight tracking data revealed that the aircraft had taken off from Shifosi Airport, northeast of Beijing, and had come close to a Hainan Airlines passenger jet, forcing at least two commercial flights to abort their landings.

For the five days following the incident, little to no official information was available on the highly unusual incident. Meanwhile, Chinese authorities reportedly ordered a nationwide suspension of most general aviation activities, including recreational flying.

Why the CITIC Tower Incident Triggered an Information Blackout

But why has the censorship surrounding this incident been so intense, leading online censors to scrub all footage and discussion of it from the internet?

This has to do with the nature of the incident.

It is not uncommon for major incidents to be heavily censored, especially when the details surrounding them are still unfolding. But this kind of information “blackout” generally occurs only when incidents are considered particularly sensitive, and the regular toolkits used to manage and control information are no longer deemed sufficient.

Some examples include:

▪️The 2011 Wenzhou Train Crash, which happened at a time when the prestige and reliability of China’s high-speed rail system was celebrated as a flagship national project.

▪️The 2013 Tiananmen SUV ‘Suicide Attack’, which happened in a tightly guarded, already politically sensitive public space and the symbolic heart of the state.

▪️The 2022 Anti-Party Protest Banner at Sitong Bridge, which happened just before the opening of the 20th Party Congress that confirmed Xi Jinping’s unprecedented third term.

▪️The 2024 Beishan Park Stabbings of American Tourists, which happened just as China’s inbound tourism was beginning to recover in the post-COVID era and “China is safe” narratives were playing an increasingly important role in official messaging.

All of these incidents are different in nature, but what they have in common is not just that they are examples of bad news being suppressed, but that they reflect a specific pattern in which news that is bad for the Party-state is suppressed. In their own ways, these incidents—and the discussions and speculation surrounding them—challenge one or more foundational narratives about the competence of China’s leadership, its modernization, its security, its stability, and its capacity to maintain social order.

📌 The 6.26 CITIC Tower incident is particularly sensitive because it touches on several of these core narratives. It immediately opened the door to a flood of politically sensitive questions: How could this happen in the capital? Were the authorities really in control? Were the security systems adequate? Was it a deliberate attack? This all directly contradicts the logic of “Safe China” (平安中国), a grand strategy for governance in the Xi era that seeks to project an image of a closely monitored society in which risks are anticipated and prevented through the integration of efficient public security, surveillance, and Party leadership (see Trevaskes & Lin 2024)¹. A small aircraft striking Beijing’s tallest building—located in one of the country’s most tightly controlled airspaces—challenges precisely this claimed capacity to maintain order, guarantee security, and keep China safe.

📌 Another reason why incidents like the 6.26 one are censored so heavily is that they have the potential to snowball or even trigger copycat behavior. What begins as an isolated accident can quickly evolve into broader discussions about political dissatisfaction, which might spiral into social unrest – and protecting domestic stability is central to the Party-state’s legitimacy and maintenance of power.

📌 A third factor is the symbolic significance of the location. Unlike an ordinary office building, the CITIC Tower is a symbol of Beijing’s cosmopolitan skyline and of China’s economic rise and state-led modernization. Situated in the heart of the capital’s central business district, next to CCTV headquarters and visible across much of the city, it is one of Beijing’s most recognizable landmarks. An incident involving such a building inevitably carries far greater symbolic weight than one occurring in a less prominent location.

⮕ Taken together, these factors make the 6.26 incident unusually sensitive in political, social, and symbolic terms. It is not just a random aviation accident; it touches upon national security, state competence, political control, social stability, and the protection of one of the country’s most symbolic urban spaces. Rather than threatening a single Party narrative, as some of the other aforementioned incidents did, it threatens several at once. This helps explain why it was treated not as an ordinary censored event, but as one requiring an almost complete information blackout.

A Familiar Censorship Playbook

Although these top-censored news incidents are all unique, they follow a similar whack-a-mole-style playbook in suppressing discussion of these events, even though they had only just happened.²

The main playbook is always more or less the same: Sensitive incident occurs → News disappears from front pages and trending search lists → Popular videos and images are taken offline → Incident-specific keywords and hashtags are scrubbed → Censorship expands to generic category-level terms (such as “Beijing,” “Chaoyang District,” or “small plane”) that could lead users back to the incident) → State media remain silent while shifting attention to other news.

(This week, the distraction might have been an otherwise arguably uninteresting controversy involving Chinese singer Han Hong.)³

Offline, the same pattern often plays out through physical scene control: bystanders are asked to delete their footage, journalists are kept away from the scene, or even entire street signs are removed (as happened after the protests on Wulumiqi Road and following the Sitong Bridge protest).

Meanwhile, selected commentary from permitted voices is sometimes allowed to appear online—for example, from former Global Times editor Hu Xijin—although even those posts are not immune from later deletion, as happened following the Beishan Park stabbings. This is often the first step in shaping the official narrative.

A brief authoritative news report may then be released, often ending with a statement that “the incident is under investigation,” which also buys time. Sometimes, however, no meaningful official explanation ever follows. After the 2022 China Eastern MU5735 crash, for example, authorities never released a full public investigation report, even though reports emerging in 2026 claimed the crash, which killed all 132 people on board, was deliberate.

Nevertheless, an official narrative and investigation outcome do usually follow. The 2013 Tiananmen attack, for example, was officially labeled a premeditated terrorist attack orchestrated by the East Turkestan Islamic Movement (ETIM), with Chinese authorities presenting it as evidence that Xinjiang-linked militancy had become part of a broader international terrorist threat. In doing so, the incident was reframed from a local security failure into part of a global fight against terrorism.

Finally, arrests often follow, and people are held accountable. At the same time, those accused of “spreading rumors” online may also face consequences, while foreign media are often criticized for biased reporting or double standards. These counterattacks further reinforce the official narrative while simultaneously shifting attention away from the original incident.

How AI Chatbots Are Part of the Censorship Ecosystem

For the 6.26 incident, the main narrative for the first five days after it happened was that nothing really happened. Even Hu Xijin posted nothing about the incident.

On Weibo, the hashtag “The Story Behind the CITIC Tower Crash Incident” (#中信集团大厦被撞事件始末#) was taken offline, along with many others.

Five days after the crash, Weibo’s Qwen-3 chatbot (powered by Alibaba) claimed it knew nothing about any incident involving CITIC Tower.

The Weibo chatbot responds to a question about the CITIC Tower incident, saying it knows nothing (July 1st).

Six days after the incident, on July 2nd, DeepSeek’s chatbot went a step further. Rather than simply claiming no information was available, it described reports about the crash as fake news, stating that there were no official records or media coverage and that such rumors originated on “overseas social media platforms.” Kimi refused to answer questions about the incident altogether, while Baidu’s ERNIE Bot said it had no information about the incident and warned users not to believe or spread unverified rumors.

[caption id="attachment_76508" align="aligncenter" width="484"] DeepSeek responded to what happened at CITIC Tower, claiming it’s “fake news” (July 2nd)

The 6.26 CITIC Tower incident illustrates a broader development in China’s online information ecosystem. A great deal has changed since the 2011 Wenzhou train crash. Social media control has become far more sophisticated—not only through better automated filtering, but also through platform regulation, agenda-setting, distraction strategies, and increasingly coordinated narrative management.

Now, AI chatbots have become part of that ecosystem.

Integrated AI assistants (LLM-powered chatbots embedded in platforms such as Weibo, Douyin, and other Chinese apps since 2025) are not merely additional tools subject to censorship. They have become an integral part of the online information environment and increasingly play a role in the handling of sensitive incidents.

Unlike Grok on X, these AI assistants are deeply integrated into the user experience and actively encourage people to ask questions, search for information, and receive summaries of ongoing events and controversies. What makes them particularly powerful from the perspective of propaganda and information control is that they can centralize official narratives at the platform level. Previously, that role was distributed across state media, government accounts, and other official actors.

Because these chatbots do not present themselves as state media or government authorities, they can appear more neutral and objective. Users may perceive them simply as AI-generated summaries with less media bias and greater factual accuracy.⁴ In the case of the 6.26 incident, they effectively conveyed a state-ordered message to social media users: that really, nothing happened, and that if something happened, it must be fake news.

“A Public Security Incident Caused by Personal Factors”

On July 2nd, Beijing Chaoyang’s verified official account issued a second short statement via its WeChat channel. This time, providing more details about what had happened on June 26.

According to the statement, the pilot was 66-year-old Mr. Liu (刘), a Beijing resident who was self-employed and lived alone. He obtained a sport pilot license in 2021 and a private pilot license in 2024.

Authorities said Liu deliberately deviated from his flight route, subsequently lost contact with the airport, and then intentionally crashed into the building, killing himself. They added that he had suffered from depression in recent years and that his diary contained multiple references to wanting to “end his life.” The incident was officially described as “a public security incident caused by personal factors” (个人原因造成的危害公共安全案件).

The second, July 2nd, statement.

It is noteworthy that such a short statement includes details about the pilot’s age, marital status, pilot licenses, mental health, and even diary entries, yet provides virtually no information about the building itself, the security response, or the broader circumstances of the incident. The statement never mentions the CITIC Tower by name, referring only to “a high-rise building near the East Third Ring Road.” It likewise omits any mention of the aircraft entering restricted airspace, the reported disruption to commercial air traffic, or the obvious questions the incident raises about security failures.

On Weibo, the statement was widely shared, but comment sections beneath those posts remained either hidden or only contained a selection of replies.

And so, six days after the incident, there is a slight opening in the media blackout—just enough to establish the official narrative. This framing reduces a wide range of sensitive questions to the story of one troubled individual. It is not about broader questions of public safety or governance, but about a lonely man’s suicide; it is not about the Party-state, but about personal circumstances; it minimizes a news event that received maximum attention and turns a possible mass attack into an isolated incident. And so the message subtly shifts from “nothing happened” to “it’s time to move on.” The playbook is complete. It’s almost as if nothing really happened.

By Manya Koetse

With contributions by Miranda Barnes

1. See: Susan Trevaskes and Delia Lin, “Integrating Stability Maintenance into Comprehensive Governance: The Burgeoning ‘Safe China’ Behemoth,” Modern China 50, no. 6 (2024): 671–701. https://doi.org/10.1177/00977004241254709.
2. Some might say the 2015 Tianjin explosions also belong here, but that disaster was so large and witnessed by so many people that, while the official narrative was tightly controlled, it did not receive the same near-total information blackout.
3. This controversy involved Han Hong (韩红), a popular Chinese singer and philanthropist, who recently urged Beijing audiences in the local dialect to show up at the box office for the underperforming latest Feng Xiaogang film Catch the Spy (《抓特务》), for which she is the music producer. She used the colloquial expression zǒu ge miànr (走个面儿, roughly “show your face”), and was accused of inappropriately putting moral pressure on people to buy tickets, which also led to criticism of her charity work. She made the remarks at a June 17 premiere, and it didn’t become a dominant story until eleven days later.
4. See also: Yan Li, “The Impact of Automated Journalism on Media Bias, Accuracy, and Public Trust: Evidence from Young Chinese News Consumers,” Humanities and Social Sciences Communications 13 (2026): 688. https://doi.org/10.1057/s41599-026-06612-6.

(follow on X, LinkedIn, or Instagram)

Spotted a mistake or want to add something? Please let us know in comments below or email us. First-time commenters, please be patient – we will have to manually approve your comment before it appears.

©2026 Eye on Digital China/Whatsonweibo. All rights reserved. Do not reproduce our content without permission – you can contact us at info@whatsonweibo.com.