China's Supreme People's Court released a 24-article opinion directing judges to use existing laws to handle civil cases involving deepfakes, voice cloning and AI recreations of the dead. The guidance prioritizes rapid remedies — including injunctions — to stop harmful synthetic content from spreading, and clarifies when platforms and users can be held liable. Officials say the approach aims to balance innovation with security and differs from both stricter European rules and looser U.S. practice.
China's Top Court Draws Legal 'Red Lines' on Deepfakes, Voice Cloning and AI 'Resurrections'
Photo Credit: iStock
China's Supreme People's Court has issued a 24-article opinion instructing judges to apply existing law to civil disputes involving deepfakes, voice cloning and AI-generated recreations of deceased people — signaling clearer, faster remedies for victims while trying to preserve technological innovation.
What the Guidance Says
The court frames the measures as defining clear legal red lines for synthetic media. Rather than creating a new AI statute, the guidance tells judges to interpret current laws to address harms tied to personal identity, reputation and privacy when AI creates an identifiable likeness or clones a person's voice without consent.
Key harms the opinion highlights include unauthorized AI-generated images or videos, cloned recognizable voices, sexualized fake content, scams, defamation and privacy breaches. Where such acts are severe, the court notes they may warrant criminal as well as civil liability.
Faster Remedies And Platform Responsibilities
One notable change empowers courts to act quickly to limit dissemination. Victims of sexualized or otherwise damaging synthetic content may be eligible for injunctions when delay could cause irreparable harm, and judges can order distribution to stop or require platforms and AI service providers to take measures to curb further spread.
The guidance clarifies liability rules: service providers are not expected to pre-screen every user prompt or model output, but they can be held liable if notified of infringing content and they fail to respond. Users who intentionally produce harmful material may also face civil — and sometimes criminal — consequences.
Broader Consumer And Market Concerns
Beyond personal-rights issues, the opinion flags consumer harms such as so-called "big-data price discrimination" — algorithms presenting different prices or terms to different customers for the same product — and fraudulent AI impersonations used in e-commerce live streams. When impersonations amount to fraud, consumers may seek punitive damages.
Wang Liming, vice-president of the China Law Society and a professor at Renmin University, said China is pursuing a path distinct from Europe and the United States: “China should avoid the constraints of both models and seek a balance between AI development and security.”
Why This Matters
The guidance acknowledges that modern AI tools can rapidly generate convincing images, audio and video at low cost, increasing the risk that both private citizens and public figures will be targeted. By enabling quicker court intervention and clarifying when platforms or users can be held responsible, the court aims to reduce harm before it spreads widely.
These developments arrive amid a widening global debate over AI oversight. Related recent incidents include a Chinese court ruling that using AI to replace workers solely to cut costs was unlawful, U.S. authorities finding AI-generated campaign emails in California, and company warnings (for example from Anthropic) that advanced models could evolve in ways that merit international caution.
Bottom line: The Supreme People's Court seeks to strike a middle course — deterring abusive synthetic-media practices and consumer harms while allowing room for AI development — and gives judges practical tools to act quickly when victims face imminent damage.
Help us improve.
