China Expands Crackdown and Sanctions in All-Out War Against Illegal AI Content

98,000 Zimeiti Accounts Sanctioned En Masse
Six-Month Crackdown Targets Advertising Market
Correction Orders Issued for Three ByteDance Apps

Chinese authorities have launched comprehensive regulations targeting the surge of illegal generative artificial intelligence (AI) content.

Chinese President Xi Jinping. Photo by EPA Yonhap News

On May 5, Yonhap News Agency, citing the Chinese state-run English-language newspaper China Daily, reported that “the Cyberspace Administration of China (CAC), the country’s internet regulatory authority, has imposed sanctions on 98,000 self-media (zimeiti) accounts that provided various news and information without disclosing their sources.” Zimeiti refers to unofficial content channels operated by individuals or small independent operators on social networking services (SNS) such as WeChat (the Chinese equivalent of KakaoTalk) and Weibo (the Chinese equivalent of X). Many of these accounts were found to have uploaded videos while concealing the fact that the content was generated by AI, in order to increase view counts.

The State Administration for Market Regulation (SAMR), China’s market regulation and supervision body, took aim at the AI-related online advertising ecosystem for the first time at the end of last month. It launched a six-month nationwide crackdown across all areas of internet advertising, targeting illegal activities including the misuse of AI.

According to consulting firm QuestMobile, China’s online advertising market was estimated to be worth 793 billion yuan (approximately 170 trillion won) last year. As of the first half of last year, it is known that three big tech companies—ByteDance, Tencent, and Alibaba—accounted for more than half of the market.

Chinese authorities view the misuse of AI and traffic abuse (view count manipulation) using automated bots as emerging risk factors, stating that “a systematic and large-scale crackdown campaign is urgently needed.” This campaign will focus on five key areas: strengthening regulatory and platform responsibilities, supervising new types of advertising, upgrading monitoring technologies, and toughening content standards. Previously, during the Lunar New Year holiday, the Cyberspace Administration of China deleted more than 708,000 pieces of AI-generated false content and other harmful materials through an online environment clean-up campaign.

Chinese big tech company ByteDance, which operates services such as TikTok and CapCut. Photo by Reuters and Yonhap News

Direct sanctions against individual companies are also continuing. On April 28, the Cyberspace Administration of China summoned ByteDance operators and ordered corrective action and imposed penalties, citing the absence of source attribution in AI-generated videos. The targeted services were three: CapCut, an AI-based video editing app; Catbox, a file-sharing platform; and Dreamina, an image-generating AI. Authorities stated that these services “violated the Cybersecurity Law, regulations related to AI-generated services, and rules on the labeling of AI-generated and synthetic content.” In response, related companies have also started to take preemptive action. Douyin, the Chinese version of TikTok, has deleted 538,000 AI videos that infringed copyright and imposed sanctions on more than 4,000 accounts since the beginning of this year.

Meanwhile, ByteDance’s video generation AI “Shidance 2.0,” which had been garnering global attention as the “second Deepfake,” postponed its global launch due to copyright controversies. Shidance 2.0 made headlines for its ability to create 15-second videos reminiscent of famous actors or popular dramas with just a few photos and short commands, but it faced backlash in the content industry over copyright issues.

This content was produced with the assistance of AI translation services.