迪丽热巴、杨紫维权：“追究到底！”

3月20日，话题#登上热搜。有网友称AI短剧《重生后，我成了娘亲的守护神》第14集开头疑似使用了演员杨紫的脸。

Chinese actress Yang Zi has condemned the unauthorized use of her likeness in an AI-generated short drama, as a related case involving Dilraba Dilmurat underscores growing legal scrutiny over AI face-swapping in China.

3月20日，杨紫工作室发文：针对某作品未经授权，利用AI换脸、深度合成等技术，擅自使用杨紫女士的肖像制作并传播视频内容的行为，严重侵犯杨紫女士合法权益。工作室已委托律师对侵权内容进行取证，并将通过法律途径追究到底！

目前相关视频已下架。

On Friday, the topic trended online after netizens alleged that a short drama used AI technology to replicate Yang's appearance in one of its episodes. Her studio said the production had used AI face-swapping and deep synthesis without authorization, seriously infringing on Yang's rights, and added that legal action is underway. The video has since been removed.

迪丽热巴被短剧AI换脸，法院判了

同日，演员迪丽热巴起诉某平台播出短剧利用AI换脸技术盗用其肖像牟利一事，也有了最新进展。

据北京互联网法院消息，法院认定，短剧制作方擅自使用深度合成技术生成与某演员高度相似的形象，侵害其肖像权；短剧播出方未尽合理审查义务，同样承担相应责任。

On the same day, a court ruling in a separate case involving Dilraba reinforced accountability for AI misuse. The Beijing Internet Court found a production company liable for infringing on an actor's portrait rights by using deep synthesis without consent. A distribution platform was also held responsible by the court for failing to conduct adequate review.

此案原告为国内某知名演员，该演员发现被告A公司制作并发布的短剧中，通过AI换脸技术将其肖像拼接至剧中角色面部，相关话题在多个社交平台引发讨论，大量网络用户误认为原告参与了该剧。与此同时，被告B公司在其运营的视频账号上线涉案短剧。

法院判决A公司、B公司分别在其运营的涉案视频账号向原告发布书面致歉声明并赔偿经济损失。双方均未上诉，一审判决已生效。

Both parties were ordered to issue public apologies and pay compensation, with the ruling taking effect after neither side filed an appeal.

部分平台开始限制“AI换脸艺人”行为

2月12日，豆包旗下视频生成模型Seedance2.0全端上线后，执行严格限制：禁止生成明星、公众人物、网红等真实人物肖像/视频/语音。

查询后豆包回复：2026年2月13日起全平台（含豆包主App）统一收紧，关键词拦截+人脸比对+内容审核三重机制，命中即拒绝生成。核心原因为肖像权、版权合规，防范AI换脸/伪造/商用侵权。限制范围包括禁止生成明星姓名、肖像、声音、视频、换脸、模仿等；限制范围不仅明星，所有未授权真实人物均受限，但用户本人数字分身（需活体校验）不受限。

The controversy comes as platforms tighten rules on AI-generated content. Following the launch of Seedance 2.0, Doubao said it would ban the generation of real individuals' likenesses, voices and videos without authorization, using filters, facial recognition and content checks to curb deepfakes and infringement.

《法治日报》律师专家库成员、北京市康达律师事务所合伙人罗娟表示，滥用AI换脸、声音克隆主要涉及民法典第一千零一十八条（肖像权保护）、第一千零二十四条（名誉权保护）。未经同意商用他人肖像、声音，直接侵犯肖像权；若生成内容虚假误导（如虚假代言、不当言论），会损害他人名誉权。

Legal experts cited provisions in China's Civil Code, noting that the unauthorized commercial use of a person's likeness or voice constitutes a violation of portrait rights, while misleading or fabricated content may also infringe on reputation rights.

来源：中国青年报 @杨紫工作室 荔枝新闻 北京晚报 中国新闻网 九派新闻 法治日报

跟着China Daily

精读英语新闻

“无痛”学英语，每天20分钟就够！