Actor Likeness Authorization / 演员脸部授权

Definition

Actor likeness authorization / 演员脸部授权 is the commercial permission to use an actor’s face, body, or identity signal in AI-generated performance without necessarily transferring the actor’s own artistic authorship to the generated output.

Current Synthesis

The Wu Hankun episode separates two questions that are often collapsed. One question is whether an actor can license a face or persona for commercial use, such as platform-generated clips or additional livestreams. The other is whether a third party controlling that face with prompts, motion, or scripts is still producing the actor’s own performance.

The current judgment is conditional: authorization can be commercially reasonable when it creates clear value and compensation, but it does not automatically solve authorship, consent scope, credit, labor replacement, platform scope, product category, or audience expectation. The actor’s face can travel farther than the actor’s intentional expression.

Platform commerce turns actor-side authorship into an enforcement problem. A celebrity or creator may authorize one account, platform, product category, or time period without authorizing AI-generated speech, altered live clips, voice substitution, or misleading product endorsement elsewhere. Platforms therefore need authorization records and appeal paths, not just a binary rule that any face-like use is either permitted or banned.

The performer-experience evidence gives another reason for keeping authorization explicit. Adah Sharma reports that digital figures can already imitate her voice and some expressions and treats consent as a live issue as those replicas become more convincing. The source does not describe a specific license or misuse case, but it strengthens the judgment that capability is not permission.

Key Claims

• A licensed face is not the same as a self-authored performance.
• Commercial authorization may still be rational when it creates value for the actor, platform, and audience.
• Consent scope matters because a third party may control script, movement, tone, usage context, account, category, and platform.
• Likeness authorization sits between IP rights, labor protection, creative authorship, and platform commerce.
• Audience and consumer trust depend on whether viewers know who controlled the performance and whether endorsement is real.
• Commerce platforms need structured authorization data because lawful likeness use may be limited by account, platform, category, duration, and synthetic-use scope.
• A performer’s recognition that imitation is technically possible does not itself authorize copying, control, distribution, or substitution.

Evidence

• Authorship split: No.232 吴汉坤：AI 时代，演员第一个丢饭碗？ records Wu’s view that a third party using an authorized face with prompts or motion may not count as the actor’s own performance.
• Commercial qualification: No.232 吴汉坤：AI 时代，演员第一个丢饭碗？ notes Wu’s concession that actors need income and that practical arrangements depend on the case.
• Livestream example: No.232 吴汉坤：AI 时代，演员第一个丢饭碗？ uses authorized streamer faces to open more livestreams as a potentially value-creating scenario.
• Industry boundary: No.232 吴汉坤：AI 时代，演员第一个丢饭碗？ distinguishes technical feasibility from whether the industry should allow substitution of workers in particular roles.
• Audience expectation: No.232 吴汉坤：AI 时代，演员第一个丢饭碗？ connects AI performers to the problem of whether viewers believe a human subject is behind what they see.
• Commerce scope: No.224 用 AI「偷」明星声音和形象侵权卖货，直播间该怎么管？ has 朱威 and 嘉旭 discuss platform, term, category, account, and deep-synthesis authorization limits in 抖音电商.
• Enforcement risk: No.224 用 AI「偷」明星声音和形象侵权卖货，直播间该怎么管？ identifies similar-looking users, model mistakes, and unsubmitted legitimate authorization as mistaken-enforcement scenarios.
• Performer consent concern: Adah Sharma on The Kerala Story, Backlash, Threats & Handling Rejection | Shekhar Natarajan | Tomorrow Today has Sharma say digital figures can imitate her voice and some expressions while explicitly raising consent and creative-employment concerns.

Counterevidence & Qualifications

The sources do not provide a contract template, union position, full platform policy, adjudicated legal test, or documented Sharma replica dispute. They define creator-side authorship, performer concern, and platform-commerce enforcement boundaries rather than a settled rights regime.

What Changed

• Added a performer’s first-person concern that rapidly improving voice and expression imitation still requires consent.
• Clarified that technical replicability or public visibility is not authorization.

Related Concepts

• AI Public Likeness Generation - broader public-likeness and deepfake boundary.
• AI Voice Cloning Rights - adjacent authorization problem for voices rather than faces.
• Synthetic Performers - generated performer category that can use or avoid real-person likeness.
• AI Content Provenance - disclosure layer needed for generated or authorized media.
• Creative Labor AI Backlash - labor and audience resistance when authorization is used to replace workers.
• Actor AI Performance Boundary / 演员与 AI 表演边界 - performance-quality and authorship boundary behind the rights issue.
• Platform Authorization Boundary / 平台授权边界 - broader contract-scope boundary for platform commerce.
• AI E-commerce Infringement Governance / AI电商侵权治理 - governance setting where likeness authorization must be enforced.