China Cracks Down on Fake Personas in Short Video

China’s top internet regulator has launched a two-month nationwide campaign to clean up short video, targeting one of the format’s most persistent problems: fabricated personas. The Cyberspace Administration of China announced the action under its “Qinglang” or “Clear and Bright” banner, a recurring series of internet-governance drives that in past years have focused on online bullying, celebrity fan culture and misleading financial content.

This edition takes aim at five categories of false identity. The first is people who pose as members of vulnerable groups, including staged clips of delivery riders hauling children along on shifts or supposedly overqualified workers taking courier jobs to dramatize hardship. The second covers invented professional credentials, such as unqualified people presenting themselves as doctors, lawyers or financial advisers.

The third and most technically complicated category involves artificial intelligence. Regulators flagged videos that use AI-synthesized audio and video to impersonate public figures, including fake endorsements and marketing built on a celebrity’s likeness without consent. A related concern is people who ride the visibility of a trending social incident by pretending to be a participant or witness.

The final category covers accounts that manufacture tragic backstories to farm sympathy and traffic, a genre Chinese viewers have nicknamed “selling misery.” The campaign notice also targets what it calls disruptive behavior by so-called internet celebrities, including staged conflicts designed to go viral.

The timing is notable. Short video is now the dominant form of online entertainment in China, with hundreds of millions of daily users across platforms such as Douyin, Kuaishou and the video arm of WeChat. Because these feeds reward emotional engagement, invented hardship stories can spread quickly and monetize through tips, livestream sales and advertising before fact-checkers catch up.

AI has raised the stakes considerably. Cheap voice-cloning and face-swapping tools mean a convincing fake of a well-known entrepreneur or actor can be produced in minutes. Regulators have spent the past two years building rules that require synthetic content to be labeled, and this campaign signals that enforcement is now moving from labeling toward active takedowns of accounts built entirely on deception.

For platforms, the practical burden is heavy. Companies will need to review accounts, verify claims about professional qualifications and respond to user reports within compressed timelines. Industry watchers expect the biggest platforms to tighten verification for accounts that present themselves as medical, legal or financial experts, and to add friction for content that leans heavily on hardship narratives.

The campaign is the latest sign that Beijing wants the short-video economy to mature rather than simply grow. The sector is a major employer of livestream hosts, editors and agencies, and advertisers have grown wary of brand safety risks around fake-persona scandals. A cleaner feed, regulators argue, is better for both consumers and the legitimate businesses that depend on the format.

Whether a two-month push can meaningfully change incentives is another question. Past Qinglang campaigns have produced visible short-term results, with account bans and featured enforcement cases. But the underlying economics, where attention converts directly into income, tend to regenerate the same tricks under new names. The coming months will show how aggressively platforms police the line between storytelling and fabrication.