AI-generated short dramas disrupt industry and violate performer rights in China
AI-generated short dramas rapidly proliferated in China in 2025–2026, dramatically reducing production costs and workforce demand in the entertainment industry. Simultaneously, multiple AI works were found to use performer likenesses without consent, while a virtual AI actor was presented with fabricated credentials, undermining both performer rights and transparency about AI systems.
What happened
AI-generated short drama series began proliferating in China in 2025 and early 2026, with works like the 81-episode series "Zhanxiántái Zhēnrén AI Bǎn" (released December 2025) reaching 100 million views in six days and "Huòqùbìng" (released February 2026) generated by a three-person team reportedly in five days at a cost stated as 3,000 yuan in computation fees. The first AIGC animated feature film "Tuányuánlìng" released 28 February 2026.
Multiple AI short dramas were discovered using AI-generated images with facial characteristics matching performer Xiaozhan without consent, including "Jīnghuá Fēngyún 1" (subsequently removed) and "Qīnghuàn Wù" (character image later replaced). Virtual AI actor Tilly Norwood was created by UK company Particle 6 and presented with falsified credentials including a stated Master's degree from Imperial College London.
Reported impact
The emergence of low-cost AI production prompted major short-drama platforms including Red Fruit to suspend or reduce real-actor projects. Industry data showed over 90% of live-action short dramas operating at loss, with major companies reporting single-quarter losses exceeding 700 million yuan. Performer compensation reportedly halved and shooting opportunities declined, with a 24-year-old scriptwriter noting that investor interest shifted away from live-action content.
Performers reported reduced work availability from early 2026, with one actress noting fewer call notices by March. Industry workers including editors, directors of photography, set designers, and especially line producers faced reduced demand, as AI-based production required minimal crew. The production model shift emphasized cost reduction over employment, with AI-generated works reportedly achievable at per-episode costs of a few thousand yuan compared to traditional productions costing hundreds of thousands.
Index analysis
The incident reflects a technological transition in entertainment production where AI systems can replace labor-intensive processes. Short-drama production had established standardized, modularized workflows (formulaic narratives, episodic hooks, factory-like script production) that aligned with AI capabilities, enabling rapid scaling of AI generation.
The misuse of performer likenesses without consent represents a failure in content moderation and rights protection, as AI systems generated images mimicking identified actors' faces. Particle 6's presentation of Tilly Norwood included fabricated credentials, indicating a broader pattern of misrepresenting AI systems' origins and capabilities.
Governance controls
Affected platforms such as Red Fruit publicly stated continued commitment to live-action content, though concurrent deployment of AI systems across iQIYI, Youku, and other platforms indicated inconsistent enforcement. No regulatory action, industry standard, or formal rights protection mechanism was documented in response to unauthorized likeness use.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.