In China, AI series are already worth 13 billion euros. If these virtual players are establishing themselves as influencers, emerging disputes over image rights are hampering this expansion.
The Evolution of Online Entertainment
The Chinese audiovisual landscape is observing the rapid development of “micro-dramas” (or playlets). These series, composed of episodes of a few minutes and designed for vertical reading on platforms like Douyin (the Chinese TikTok), constitute a new consumption format.
The most notable development in 2026 lies in the mode of production: these formats now rely largely onartificial intelligence generative. By reducing logistics costs and accelerating the pace of creation, this segment now represents nearly 12 billion euros. The recent launch of ‘Master of Feng Shui’ (100 million views in 12 hours for its first episode) illustrates the adoption of this format by the public.
From avatar to intellectual property: the quest for authenticity
The use of on-screen AI is part of a gradual adoption curve:
Qin Lingyue (秦凌岳) and Lin Xiyan (林汐颜) (Weibo)
This approach marks a change in strategy: these characters are developed as independent intellectual properties (IP). With their own accounts on Douyin and RedNote (the Chinese counterpart of Instagram), they aim to cultivate the “feeling of being alive” or perceived authenticity. The objective for the studios is to create virtual influencers capable of engaging audiences outside of fiction, while maintaining total control over their public image.
A real post from AI actress Qin Lingyue’s RedNote account (Weibo)
The legal challenges of “digital doubling”
However, this economic model raises new legal questions. To achieve this level of realism, AI models train on existing data, raising the thorny issue of consent and intellectual property.
Debates recently emerged when the features of virtual actors Qin Lingyue and Lin Xiyan were perceived to have strong similarities with real Chinese actors (Zhai Zilu, Zhao Jinmai and Zhang Zifeng). Faced with these practices, the courts are beginning to establish initial case law. In March 2026, Beijing courts sanctioned two companies for having transposed the features of an actress onto a virtual character without authorization.
The legal framework is tending to become clearer: even if the generated face presents slight modifications, the offense can be characterized as long as the public recognizes the original person. Broadcast platforms are also invited to strengthen their moderation under penalty of incurring liability.
Is the AI influencer model exportable to Europe?
The Chinese micro-drama industry constitutes an interesting case study for European marketing departments. It demonstrates that generative video makes it possible to produce narrative content on a large scale and to consider the creation of virtual brand ambassadors.
However, a direct transposition of this model in the West would come up against two major limitations. The first is legal: the question of image rights requires ensuring that models are based on strictly royalty-free data, in order to avoid complex and prejudicial disputes.
The second limitation is cultural. It is indeed appropriate to question the reception of these AI actors by the European public. If the Chinese audience appears more and more receptive to this hyper-virtualization, Western consumers could express a certain distance. The attachment to human authenticity, imperfection and the true emotional incarnation of an actor remains a strong marker in Europe. Beyond economic or legal viability, the central question for brands will be whether the European public is truly willing to place their trust in an algorithm-generated entity, or whether they will continue to favor connection with real actors.
