In China, a fundamentally new market for licensing biometric data has formed, where ordinary citizens, former actors, extras, and influencers can legally rent out images of their faces for use in AI-generated content. This segment is already actively used in creating micro-dramas, advertisements, and video games. Specialized digital platforms offer producers extensive catalogs of appearances, and project creators actively select suitable types for their tasks.
Services like ActID, based in Shenzhen and launched in March, and New Claw from Chengdu provide users with tools to independently define the permitted spheres of application for their digital doubles. Rights holders also set the cost of using their data. The price range varies from $15 to $700 per series or a single image. The ActID platform has about 800 registered users, of which roughly 300 have given official consent to license their biometric data. Two AI projects have already taken about 10 faces into work for content generation. The platform's commission is 10%. To avoid price dumping and devaluation of labor, New Claw set a minimum price of $74 per image.
Operational director of New Claw, Long Lyu, points out that selling rights to photographs is becoming a reliable way to earn income while continuing a traditional offline career. Marketing director of ActID, Camille Yan, notes that producers are looking for both attractive faces for lead roles and specific characteristic types, such as elderly people. According to her, the industry needs living human bases, as fully synthetic faces look unnatural. In AI dramas, acting skills are not required, as faces serve solely as source material.
The scale of the emerging industry is impressive. According to data from the China Netcasting Services Association, published in specialized publications in May 2026, about 128,000 micro-dramas were released nationwide in the first quarter of 2026. Of this total, AI micro-dramas accounted for approximately 122,000, exceeding 95% of all released projects. The association regularly publishes official statistics, confirming a steady shift towards automated video content production.
Key Features of the Market
Key features of the emerging segment include:
Detailed filtering of catalogs by age, gender, and specific type (e.g., "girl next door," "rugged," or "supermodel") depending on the project's genre;
The ability for rights holders to independently and strictly limit the spheres of use for their digital image;
A complete lack of need for traditional acting skills among biometric data providers.
Official websites of the ActID and New Claw platforms, as well as their press releases or publications on social network X, were not found in open access. Information about the companies' activities is based on direct interviews with representatives by the publication Rest of World. No more recent data on the platforms themselves from recent days was found, but the general vector of the industry's development is obvious.
The development of platforms for licensing digital twins indicates a deep structural transformation of the entertainment industry in China. Automating visual content creation reduces producers' dependence on traditional casting, physical filming, and logistical constraints.
AI Opinion
Analysis shows that the demand for real biometric data is directly related to the "uncanny valley" effect. Fully synthetic faces often cause subconscious rejection in viewers due to unnatural facial expressions and lifeless skin, which reduces engagement. Using real photographs as a base allows neural networks to preserve micro-expressions that algorithms cannot yet generate from scratch. The situation demonstrates a shift from pure generation to hybrid modeling, where humans act as living templates.
A historical parallel is the stock photography market of the early 2000s, but now digital casts of biometrics are entering circulation. This creates new vulnerabilities. Researchers have long warned about the risks of unauthorized cloning of appearances beyond licensing agreements. Digital rights protection technologies still lag behind the speed of development of generative models. How will the industry balance commercial benefit and personal data security in the long term?
▼
The most interesting and important news on our Telegram channel
end-content
