The Cyberspace Administration of China (CAC) recently issued a notice announcing a three-month nationwide campaign titled "Clear and Clear: Rectifying the Abuse of AI Technology." According to information released on its official website on April 30th, the campaign will be conducted in two phases, with a primary focus on combating the "use of AI to fabricate and disseminate rumors."
Former Sina Weibo censor: Marketing accounts are the main target
According to a release from the Cyberspace Administration of China (CPC) Central Committee, "using AI to fabricate and disseminate rumors" includes "creating all kinds of rumors and information related to current politics, public policy, social livelihoods, international relations, and emergencies, or making arbitrary and malicious assumptions about major policies." Furthermore, there are also "making up or fabricating causes, progress, and details of emergencies and disasters, impersonating official press conferences or news reports, and spreading rumors. Furthermore, content generated by AI's cognitive biases is used to maliciously mislead people."
Liu Lipeng, a former censor at Sina Weibo, believes that in China's current political climate, voices opposing the authorities are becoming increasingly difficult to maintain. The Cyberspace Administration of China's statement about "using AI to fabricate and spread rumors" is primarily aimed at marketing accounts. "This targeting targets marketing MCNs (multi-channel networks). 'Anti-rebel' speech has likely long since crossed their bottom line, and whatever is said falls within their crackdown."
China's official rumor-mongering regulations
Liu Lipeng told this station that with Chinese public opinion currently saturated with the authorities' nationalist propaganda, many marketing accounts are using AI to independently fabricate highly exaggerated nationalist content, generating a large number of articles. "This is counterproductive to their (Chinese authorities') propaganda efforts. The AI's fabrications may conform to the exaggerated patriotic propaganda in some aspects, but they also fabricate many non-existent details, making them excessively exaggerated, and the output is extremely large."
The Taiwan Democracy Lab, a Taiwanese non-governmental group, published analysis results last year, showing that many China-related accounts used new technologies and AI to generate rumors, thereby exacerbating internal public opinion disputes and conflicts in Taiwan.
Liu Lipeng stated that these targeted rumors are directly instigated by Chinese officials, unlike accounts that independently promote exaggerated nationalist content. These rumors directly instigated by Chinese officials have clear copywriting and are likely generated by online trolls commissioned by Chinese officials.
He explained the criteria used by Chinese officials to create this type of targeted false content: "Generally, these accounts follow the official ones, and some have some communication with the official, who then ask them to post certain things. Then, there are some mid-level and lower-tier patriotic marketing accounts that simply follow behind these patriotic traffic. Even the slightest misstatement in the copywriting would constitute a violation, and the responsible individuals could be arrested."
We-media practitioners worry about further tightening of China's speech environment
The Cyberspace Administration of China (CAC) and the Cyberspace Administration of China (CAC) are administrative offices of the CPC Central Committee and the State Council, respectively, representing "one agency, two names." Operation Qinglang (Clear and Bright) is an annual internet campaign launched by this agency since 2016, with a different focus each year. In the 2016 Operation Qinglang (Clear and Bright) alone, Chinese authorities announced the arrest of over 17,000 criminal suspects.
An interviewee with years of experience running self-media platforms in China, speaking anonymously for security reasons, pointed out that the Cyberspace Administration of China's Central Committee's "Crackdown on the Abuse of AI Technology" campaign will likely lead to further censorship of speech. He said, "There's actually no clear definition of what constitutes 'fabrication' or 'malicious interpretation' when it comes to current political issues. Sincerely commenting on current affairs, even if your tone differs from the government's, can lead to accusations of this. This only makes people more fearful and afraid to discuss current political issues."
