entertainment 5 min read

China's AI Drama Factory Is Eating the World

China produced 128,000 short dramas last quarter — 95% AI-made, at 10% the cost. What starts as a Chinese labor shock could become a global flooding of content that upends Korean and Hollywood production.

  • Korean Entertainment
  • Chinese Short Drama
  • AI Actors
  • Content Production
  • Streaming Industry

The Math Behind the Shock

In the first quarter of 2026, China released approximately 128,000 short dramas — three times the total the industry had produced in all of the previous year. Of those, 95 percent were made with AI. A single Qinghua University research estimate puts the production cost at 600 to 800 yuan per minute, roughly one-tenth of what human-led shoots cost. One AI animation studio alone, Ji Lin, churned out about 8,000 one-minute episodes in May.

That last number deserves a moment: eight thousand pieces of scripted drama in a single month, made by a company whose workforce presumably fits in a handful of rooms. This is not incremental change. This is a structural rupture.

The technology driving this is ByteDance’s SiDance 2.0 and similar models that allow a single operator to produce a coherent short drama in one to two days — tasks that formerly required a team of five people working for three months. The prompt is no longer speculative. It is already on screen.

Who Got Fired

Greg Werner, a Beijing-based short drama actor and producer, told the Financial Times he was shooting three shows a week before SiDance 2.0 launched. Afterward, his work effectively disappeared. Other actors around him shifted to different industries entirely. The Chinese short drama sector employed roughly 690,000 people directly before this pivot. The broader livestreaming economy adds another 15 million. These are not abstract statistics — they are livelihoods.

The labor disputes are already appearing in Chinese courts. In Hangzhou, a worker sued after being replaced by an AI replica and won additional compensation from the company. Another case involved a children’s education program host who was told by her employer: sign the rights to your AI clone, or we’ll find someone who will. She negotiated a five-year limit on usage and secured a one-year salary plus a veto right. Not every actor gets that outcome.

The so-called distillation problem — where performers’ faces, voices, and acting styles are harvested into AI models without consent — has become a new front in the fight over likeness rights across China.

The Livestream Commerce Precedent

Before drama, live commerce felt the shock first. JD.com reported that by March, more than 70,000 sellers were using AI digital hosts. These avatars work 24 hours a day for roughly 10 percent of a human anchor’s cost. Luo Yonghao’s AI duplicate, in some product categories, outsold the real Luo — recording 55 million yuan in sales over seven hours at one point.

This pattern matters because it shows the economic logic in motion. When an AI host can generate more revenue than the original at a fraction of the cost, no rational producer goes back to the human. The choice is not cultural. It is arithmetic.

The Korean Angle

Korea watches closely, and for good reason. The Korean drama and variety industry depends on a dense network of supporting actors, crew, and on-camera talent — a ecosystem that sits precisely where Chinese AI production is now aimed. Many of the actors losing work in China occupy the same middle tier that feeds Korean content: reliable professionals who appear across dozens of productions and sustain entire crews between gigs.

K-drama has competed with Chinese web dramas on affordability, speed, and cultural proximity. That margin was never generous. With Chinese short drama production costs collapsing to 10 percent of their former levels, the entire comparison shifts. A Korean production that once justified itself by being faster and cheaper than Hollywood now faces competition from something that is faster and cheaper than itself.

The question for Korea is whether its content industry can maintain its cost and velocity advantage, or whether the economic floor has moved beneath it.

The Global Streaming Implication

Western outlets have not yet absorbed the full implication of what is happening in China. The assumption that AI-generated drama will remain a domestic curiosity is already wrong. The format — vertical, episodic, serialized micro-drama — is the same format flooding TikTok, Instagram Reels, and YouTube Shorts globally. The distribution channel is borderless. The production cost is measured in minutes, not months.

Netflix, Disney+, Amazon, and every regional streamer are betting on high-budget scripted content. That bet is not wrong — it is simply operating in a different tier than the content now being produced at ten percent of the cost. But tier-two and tier-three content is where the volume lives, and where the audience spends most of its time.

If Chinese AI drama enters global streaming markets at scale, it will not compete with The Crown. It will compete with the shelf space below it — the long tail of content that fills platforms and drives engagement. That tail is enormous, and it is now being produced by machines instead of people.

What Comes Next

Several outcomes are plausible, and none of them are mutually exclusive.

First, a licensing cascade. AI companies may begin offering licensed likeness packages at scale — performers monetize their digital clones rather than fight them. The Chinese educator who negotiated a five-year limit represents one path. More performers will face the choice between participation and irrelevance.

Second, a regulatory response. China has shown it can move quickly to restrict AI-generated content, particularly around likeness rights and labor disputes. The Hangzhou court ruling suggests the legal system is already adapting. Korea and the United States are likely to follow with their own frameworks.

Third, an output glut. At current trajectories, the volume of AI-produced drama will outstrip human consumption capacity. The real constraint will shift from production to curation — which pieces surface, which algorithms promote them, and which human audiences find a reason to stop scrolling.

The Chinese short drama industry employed 690,000 people. That number is already shrinking. What replaces it is not yet clear — but the arithmetic is undeniable, and it is no longer confined to China.