Back to News
PolicyAI Understanding briefing

China’s top court sets liability rules for AI deepfakes and hallucinations

Startup Fortune reports that China’s Supreme People’s Court issued nationwide judicial guidelines addressing AI deepfakes, voice cloning, chatbot hallucinations and algorithmic price discrimination. The report’s details have not been independently confirmed from the court’s published text.

4 min readRead the primary source
Source-provided image accompanying China’s top court sets liability rules for AI deepfakes and hallucinations
Source referenceSource recorded
Publisher
startupfortune.com
Source link
startupfortune.comhttps://startupfortune.com/chinas-supreme-court-sets-first-legal-rules-for-ai-deepfakes-and-lies/
Source type
Linked source — primary-source status has not been established.

Story last revised

ContextUnderstand this in 60 seconds

Start here

Key terms

Generative AI
AI systems that produce new content such as text, images, audio, video, or code.
Hallucination
When a model generates fluent but false or unsupported information.
Prompt
The input instructions and context provided to a generative model.
Test yourselfAI Ethics Quiz

What happened

Startup Fortune reports that China’s Supreme People’s Court issued 24 provisions on Sept. 7 covering civil disputes involving AI-generated content and algorithmic decisions. The reported framework can assign responsibility to creators, platforms and users, including when flagged harmful content remains online.

Startup Fortune reports that the Supreme People’s Court issued its first nationwide judicial guidelines on artificial-intelligence disputes on Sept. 7. According to the report, the rules address AI deepfakes, cloned voices and faces, chatbot hallucinations, defamatory falsehoods and algorithmic price discrimination. The report says people who create harmful fakes, platforms that host them and users who deliberately prompt them can each face liability in relevant circumstances.

The report says a provider is not automatically liable for every hallucination, but may share responsibility after receiving notice of a false and damaging output and failing to act. It also says deliberately prompting a model to generate defamatory or fabricated content would not shield a user from responsibility. Startup Fortune cites a reported case in which a finance worker was deceived into transferring 1.86 million yuan after a fake video call, and a separate Hong Kong police account involving an approximately $26 million loss. Those examples and the legal interpretation have not been independently confirmed here. No access process or pricing applies to these judicial guidelines, and the source does not provide the court’s full text.

Source details: startupfortune.com

Why it matters

The reported guidelines could give Chinese courts a more specific basis for assigning responsibility when generative AI causes reputational, financial or consumer harm. Their importance depends on the official text, judicial interpretation and enforcement across local courts, none of which Startup Fortune’s report independently establishes.

If accurately reported, the framework would make platform response procedures part of AI risk management in China. Companies operating consumer-facing generative AI services may need to identify, review and remove flagged deepfakes or harmful hallucinations, while users could face exposure for intentionally soliciting deceptive or defamatory material. That could affect product moderation, evidence preservation and complaint handling.

The practical significance should not be overstated. Startup Fortune’s comparison with the European Union and United States is the outlet’s characterization, not an independently established global legal ranking. The report does not establish how the provisions interact with existing Chinese law, how damages would be calculated, or whether courts will apply the standards consistently.

What to watch next

The court’s full published opinion, early cases applying it, and how platforms implement notice-and-action procedures will determine the rules’ practical reach. It remains unclear whether the framework creates new liability beyond existing Chinese civil, consumer-protection and platform obligations.

The first priority is verification against the Supreme People’s Court’s official publication and authoritative Chinese legal commentary. Key questions include the precise definitions of deepfakes, hallucinations, notice, intent and platform responsibility, as well as whether the provisions are binding rules, judicial guidance or interpretive principles.

Future cases will show whether liability attaches to model providers, hosting platforms, content creators or prompt users in different combinations. Watch for enforcement involving voice and face cloning, fraud, defamation and personalized pricing, along with evidence about takedown response times and remedies for victims. Startup Fortune’s report does not establish implementation dates, penalties or an appeal process.

Related guides & quizzes

AI EthicsChatGPT & LLMsAI Models ExplainedTest what you know — try a free AI quizLook up an AI term in our glossary
Found this useful?