China’s broadcasting regulator has set out a policy preference for live-action microdramas while placing new visibility and rights-protection expectations around AI-produced titles. On September 17, Wang Xiaoliang, director general of the National Radio and Television Administration’s Network Audio-visual Program Management Department, addressed the sector’s next five years at a press conference. As South China Morning Post reported on September 30, Wang linked the rapid rise of AI microdramas to questions of content quality and rights protection, and said the NRTA should strongly support live-action dramas as the main source of high-quality production.
That direction sits alongside a formal management framework that took effect September 1. Wang’s preference for live-action productions is a separate policy position, rather than a provision of Order No. 16. In the same public discussion, he described AI as a creative tool with potential to reduce costs and improve efficiency. Separately, official reporting said platforms will be required to verify that creators have permission to use a person’s likeness or voice and to strengthen checks during production and distribution.
Wang Xiaoliang’s Live-Action Priority for Microdramas
The scale cited by Wang helps explain why the NRTA is specifying that balance. Official data cited by Wang put January-through-August microdrama releases at roughly 430,000. He said that was 13-fold the number released across all of 2025. Wang also said more than 90% of those releases used AI technologies. Those are official figures rather than an independently audited census of either releases or production methods, yet they describe an output volume that has made platform rules consequential to how audiences encounter the format.
For Wang, the policy response includes active support for live-action titles. Official reporting says platforms are encouraged to create dedicated live-action sections and offer financial support so well-produced shows can reach audiences and generate returns. That is an encouragement and a stated policy preference, rather than evidence that every platform has created such a section or that live-action production has become mandatory. It nevertheless makes a quality hierarchy explicit in a field where AI use has become common in officially reported releases.
The same official coverage places microdramas within China’s 15th Five-Year Plan for 2026 through 2030 as an emerging cultural-business form to be actively developed. The NRTA has also stated a goal of releasing 1,000 high-quality microdramas in 2026, focused on contemporary life, regional culture, and international communication. Both points describe regulatory ambitions and targets, rather than completed results. The distinction matters because the regulator is seeking more production in the format while defining standards for the technology used to make it.
Earlier EastFrontier coverage of China’s AI-driven microdrama factory documented the production-scale context. Wang’s September intervention concerns a different lever: what platforms should elevate as high-quality work, and what checks should surround AI-made releases once they enter distribution.
Order No. 16 Makes AI Disclosure a Per-Episode Requirement
The binding foundation is the NRTA’s Microdrama Development and Management Measures, Order No. 16. The measures were approved July 27, published July 31, and effective September 1. They define a microdrama as a serial work with individual episodes under 20 minutes. They also direct creation and broadcasting to respect originality, protect copyright, and develop in an orderly manner.
Article 34 addresses AI directly. A microdrama generated or produced with AI must comply with applicable national rules and carry a prominent notice in every episode. The requirement is consequentially specific: disclosure travels with each episode, rather than being confined to a production page or a platform-level notice. It does not declare AI microdramas off limits. Instead, it establishes an audience-facing condition for works made with the technology.
The measures also contain provisions beyond AI, including categories based on investment amount and subject matter, plus filing, public-notice, review, or distribution-authorization requirements for specified types of microdrama. Treating Order No. 16 as an AI-only rule would miss its broader scope. Its AI notice provision is one element in a framework for a short-form serial category that the NRTA is regulating as such.
Official accounts add a platform responsibility that goes beyond a label. The State Council Information Office said platforms will be required to verify that creators have permission to use a person’s likeness or voice and to strengthen checks during production and distribution. That direction does not identify an individual misuse case or spell out a single consent process. It does place authorization verification alongside content review as a stated regulatory expectation.
The prominence of per-episode notices distinguishes this intervention from the earlier question of where an AI drama could appear. EastFrontier’s account of AI drama’s prime-time test in China addressed a distribution and programming development. Order No. 16 requires the notices, while official reporting separately describes planned platform checks on permissions and on production and distribution.
Enforcement Data Puts Platform Review at the Center
Wang’s enforcement figures show why the NRTA has coupled its development plans with platform oversight. According to official reporting, authorities had removed 68,000 microdramas since the start of 2026 and directed platforms to act against more than 1,200 accounts. Wang said more than 90% of content covered by those actions involved AI-produced microdramas. The totals should be read as official enforcement figures, not as a measurement of all problematic AI content.
The actions covered vulgar or harmful material, content unsuitable for children, and copyright violations, according to the SCIO account. They should not be recast as a count of copyright cases alone. Nor do the reported actions establish that a particular performer’s likeness or voice was used without permission. The policy line is preventative as well as punitive: official reporting says platforms will be required to check authorization and content through production and distribution, while Order No. 16 requires AI-made episodes to display their status.
That focus on rights protections touches concerns adjacent to, but separate from, EastFrontier’s earlier examination of actors losing work as AI microdramas spread. Wang’s stated measures do not settle questions about employment, compensation, or any individual creator’s claim. They put traceability, authorization, and platform review inside the regulator’s response to AI production at official scale.
The emerging line is therefore concrete. Live-action dramas receive the NRTA’s stated support as a quality-production priority. AI-made microdramas remain part of a format that the government intends to develop, subject to a prominent notice in every episode under Order No. 16, copyright protections, and enforcement when content breaches applicable rules. Official reporting also describes a platform direction on permission checks for likenesses and voices. How consistently platforms implement those requirements, and how the NRTA measures compliance, are questions the official materials do not answer. They establish the conditions and policy preference, rather than a final account of their results.