China's regulator sets three principles on AI likeness and voice in microdrama
The NRTA convened 17 platforms and producers on 2 September to agree that any fused likeness or voice requires the rightholder's express authorisation, and that platforms carry a review duty.
What this basis means: Regulatory filing, government body, court or industry association.
China's National Radio and Television Administration (NRTA) convened a symposium on intellectual property and personality rights in the use of AI-generated likeness and voice, publishing an account on 4 September 2026. The meeting, held on 2 September, was organised by the regulator's Department of Online Audiovisual Programme Management and hosted by the China Netcasting Services Association.
The published account records three points of consensus. First, any use of fused likeness (融像) or fused voice (融声) requires the express authorisation of the rightholder. Second, infringement is to be assessed on the overall features of the generated content, the scenario of its use, the subjective intent of the user, and whether the platform discharged its review duty. Third, participants agreed to develop end-to-end risk controls: authorisation management before production, technical screening combined with human re-review during it, and traceable evidence retention afterwards.
Seventeen platforms and production companies attended, among them Dianzhong, Hongguo (红果), Jiuzhou, Kuaishou, Mango TV and China Literature, alongside iQIYI, Tencent and Youku.
The symposium follows NRTA Order No. 16, the Administrative Measures for the Development of Microdrama (微短剧发展管理办法), which took effect on 1 September 2026 and requires AI-generated microdrama to carry a visible label on every episode.
Companies Hongguo, Kuaishou, China Literature
Tracked instrument China, NRTA Order No. 16 — Measures for the Administration of Micro Short Drama Development (微短剧发展管理办法)
Sources
- 网络视听节目管理司组织召开微短剧融像融声知识产权与人格权益座谈会 (opens in a new tab) National Radio and Television Administration