The Deepfake Dilemma: How China's Streaming Giants Are Redrawing the Line on AI and Actor Identity

iQiYi's launch of a licensed AI avatar database for Chinese filmmakers raises urgent questions about consent, creative labor, and the future of screen performance — questions that Hollywood's own debates have yet to resolve.

By Monexus Staff Writerchina-east-asia5-minute read25 Apr 2026☆ Save ↗ Share ⎙ Print

On paper, it sounds like a straightforward efficiency play: Chinese streaming giant iQiYi has built a licensed database of AI-generated avatars modeled on real, consented actors, available for deployment in upcoming film and series productions. The pitch is lower production costs, faster shooting schedules, and the ability to keep aging stars in frame without the physical toll of traditional production. In practice, the launch has exposed a fault line that the global entertainment industry has been circling for years — one that Beijing is now forced to navigate at speed.

The debate playing out in Chinese industry circles and on social platforms this week is not fundamentally different from the one that consumed Hollywood after the 2023 SAG-AFTRA strikes, when AI-generated likeness rights became a central negotiating flashpoint. But China is moving down a distinct path: rather than waiting for litigation to establish norms or for guild protections to crystallize, iQiYi has moved to commercialize the technology within a framework the company frames as consensual and structured. Whether that framework holds up to scrutiny is another matter entirely.

Consent and the Terms of the Deal

The core question is what "consented" actually means in this context. iQiYi's model requires actors to grant rights to use their likeness in AI-generated form — a transaction that, in most Western legal frameworks, would require clear articulation of scope, duration, and compensation structures. Chinese entertainment law does not yet have a dedicated statutory framework governing AI-generated likeness rights, which means the contractual terms are being set by individual platforms and producers, often with leverage asymmetry baked in.

Industry observers in Beijing have noted that the standard agreements being circulated by major streaming platforms give rights holders limited ability to control how their AI avatars are used across different productions, genres, or commercial applications. An actor who consents to appearing as a younger version of themselves in a period drama might find their avatar subsequently deployed in promotional materials, merchandise tie-ins, or even formats they find objectionable — all without further consultation, provided the original contract's fine print permits it.

This is not a hypothetical concern. Following iQiYi's announcement, several Chinese actors' representatives began quietly querying whether existing contracts signed prior to the AI avatar push could be read as implied consent to likeness digitization. The answer, in most cases, is unclear — and the ambiguity benefits the platform, not the performer.

The Creative Labor Problem

Beyond individual consent, there is a structural dimension to the debate that gets less attention in the Western press coverage. China's film and television industry is mid-consolidation: streaming platforms have been losing money at significant scale, subscriber growth has plateaued, and the cost of star talent has become a persistent margin pressure. AI avatar technology promises a partial solution — not to replace actors entirely, but to extend the commercial life of their likenesses in ways that decouple performance income from physical production demands.

The concern among mid-tier Chinese actors and production houses is that this technology, if widely adopted, will hollow out the tier of roles that traditionally serve as career-building ground for emerging talent. If established stars can be digitally sustained across productions indefinitely, and their avatars can perform younger or action-intensive versions of characters at lower cost, the pipeline that feeds the next generation of Chinese screen performers narrows considerably. This is a labor market problem with cultural downstream effects: fewer entry points means fewer voices breaking through, which over time affects the range and diversity of Chinese cinema.

Competing Approaches to Regulation

The iQiYi launch arrives at a moment when Beijing is actively developing its AI governance posture. The Cyberspace Administration of China issued guidelines in 2023 on generative AI that included provisions on intellectual property and deepfake content, but those guidelines were written before commercial AI likeness services were fully operational. Regulatory updates are expected, but their timing and scope remain uncertain.

This creates an interesting contrast with the Hollywood approach, where SAG-AFTRA's 2023 strike produced a tentative agreement that set baseline protections — including consent requirements, compensation floors, and restrictions on digital recreation — that the union has since been working to enforce through collective bargaining and, where necessary, litigation. China's streaming platforms are operating in a more permissive contractual environment, but one that carries different long-term risks: if actor communities lose confidence in the fairness of AI licensing terms, the talent pipeline that platforms depend on for premium content could become adversarial.

Some Chinese studios have signaled preference for industry-level agreements that would standardize avatar licensing terms, creating a clearinghouse model similar in concept to music licensing societies. Others are moving faster, cutting individual deals with high-profile talent and establishing facts on the ground before any regulatory framework constrains their options. iQiYi appears to be in the latter camp.

The Stakes for Global Standards

What makes the iQiYi story significant beyond China's borders is the question of norms. The entertainment industry has never had a truly global standard for performer likeness rights — different jurisdictions operate under different legal traditions, and the digital reproduction of human appearance has outpaced every attempt at harmonization. China's move toward commercial AI avatar services at scale, without waiting for international consensus, effectively sets a de facto standard for markets where regulatory frameworks are still catching up.

If Chinese streaming platforms successfully monetize AI-actor likenesses and demonstrate commercial viability, that model will be examined closely by studios in India, Southeast Asia, and the Middle East — markets where entertainment production is growing rapidly and where regulatory capacity is often limited. The question is not whether AI-generated actor likenesses will become a commercial fixture — they almost certainly will — but whether the frameworks governing that fixture will be actor-protective or platform-dominant.

For now, the answer in China depends on how courts interpret existing contracts, how regulators choose to act, and how the actor community organizes its response. What is clear is that iQiYi's move has accelerated a conversation that Beijing hoped to conduct on a slower timeline. The deepfake dilemma, it turns out, arrives on its own schedule.

This publication covered the iQiYi avatar launch as a technology and rights story; wire coverage in Western outlets focused primarily on the novelty angle, with less attention to the structural labor implications that Chinese industry sources identified as the central concern.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

https://t.me/france24_fr/31428