China released this Monday (7) a set of judicial guidelines for the country to judge disputes involving Artificial Intelligence (AI). Announced by the Supreme People’s Court of China, it provides for liability for deepfakesvoice cloning and unauthorized use of third party images.
The document establishes consent as the most important element in defining whether the use of AI should be considered legal or illegal. According to information from the Xinhua agency, the guidelines “make clear that people cannot use AI to distribute recognizable digital replicas of other people without their consent, including cloned faces and voices”.
The new standards also provide for liability for AI service providers, who may face legal liability if they do not act quickly after being notified that their systems generated content that violates the rights of others.
According to the vice president of the court, Tao Kaiyan, the guidelines seek to balance development and security, clarifying procedural rules for AI litigation, while leaving open the points on which there is still no defined consensus on the topic.
In April 2024, the Chinese judiciary faced, for the first time, an AI-generated voice infringement lawsuit. At that time, it was decided that recreating a person’s identifiable voice without their consent would be illegal. According to information from Xinhua, the Guangzhou Court alone has judged, in the last three years, around 700 cases of “face theft” by AI.
Over the past few years, China has sought to strengthen its legal framework on the topic. In 2023, for example, the “deep synthesis” regulation came into force, aimed at content generated or altered by AI. Meanwhile, in 2025, rules requiring labeling of AI-generated materials were published.
Last week, China said it removed more than 5.6 million pieces of harmful content and about 49,000 accounts in a campaign against AI misinformation on platforms including WeChat, Douyin and Kuaishou.
Source: www.brasildefato.com.br
