# China's Supreme Court Sets First Legal Rules for AI Deepfakes and Lies > Source: > Published: 2026-09-08 02:34:56+00:00 *China's Supreme People's Court has drawn the world's first detailed legal line on AI harm, deciding exactly who pays when a chatbot lies or a deepfake scams someone.* China's Supreme People's Court issued its first nationwide judicial guidelines on artificial intelligence disputes on September 7, 2026. That's new. For the first time anywhere, a major economy has spelled out exactly who owes what when an AI system lies, clones a voice, or fakes a face. The rules cover deepfakes, chatbot hallucinations, and algorithmic price gouging. Three parties can be held responsible: the person who made the fake, the platform that hosted it, and the user who typed the prompt. The guidelines, formally titled the Opinion on Properly Hearing Cases Involving Artificial Intelligence Disputes, treat a cloned face or voice as theft the moment someone builds it without consent. Simple as that. You do not get to build a talking avatar out of someone else's face and voice just because the tools make it easy. Do that to spread a false claim or damage someone's reputation, and you're on the hook twice. Once for the copy. Once for what you did with it. Then there is the harder problem. What happens when an AI system just makes something up? Generative models hallucinate constantly, inventing quotes, fabricating details, describing events that never happened. Under the new rules, the company running that model is not automatically liable for a hallucination it did not know about. But once someone flags the false, damaging output and the provider sits on it, the clock starts. Fail to act, and the platform shares the blame with whoever asked the question in the first place. That second half matters more than it sounds. A user who deliberately prompts a model to produce something defamatory or fake, knowing it is false, cannot hide behind "the AI said it." Zhou Jiahai, head of the Supreme People's Court's research office, put the reasoning plainly: "We cannot expect every consumer to become an expert at spotting deception. The law must step in promptly to protect consumers' legitimate rights and interests." [Arm CEO Rene Haas Says Chip Shortages Are Slowing Down AI Cancer Research](https://startupfortune.com/arm-ceo-rene-haas-says-chip-shortages-are-slowing-down-ai-cancer-research/) Arm CEO Rene Haas told the BBC that AI could eventually crack cancer by modeling how DNA markers drive tumors, but said the breakthrough is being held back by a shortage of chips needed to build AI data centers. He also predicted humanoid robots doing repair and manufacturing work within a decade and downplayed fears of mass AI job losses. - [chip shortage slowing down AI cancer research breakthroughs](https://startupfortune.com/arm-ceo-rene-haas-says-chip-shortages-are-slowing-down-ai-cancer-research/) - [how chip shortages are impacting AI medical research](https://startupfortune.com/arm-ceo-rene-haas-says-chip-shortages-are-slowing-down-ai-cancer-research/) The timing is not a coincidence. Deepfake-driven fraud has surged across China this year, part of a wider global pattern. A finance worker in China was tricked into wiring 1.86 million yuan, about $262,000, after a video call with what looked and sounded exactly like her boss. It was not her boss. It was a clone built from public video and a cloned voice track. A similar scam cost a Hong Kong-based multinational roughly $26 million in a single incident involving a faked video conference call, according to Hong Kong police. Cases like these are exactly why judges needed rules instead of guesswork. The court also went after a quieter harm: algorithmic price discrimination, the practice sometimes called "big data killing familiarity" in China, where a platform charges loyal or frequent customers more than new ones for the identical product. The guidelines say merchants who use algorithms to set different prices or terms for the same good, without reasonable justification, can be held liable for the difference. It is a separate problem from deepfakes. It comes from the same instinct: do not let the algorithm decide who gets cheated. Vice President Tao Kaiyuan framed the guidelines as an effort to balance development and security, leaving room for interpretation as new kinds of AI harm show up that nobody has anticipated yet. That is a deliberately loose phrase. It has to be. Twenty-four provisions cannot anticipate every way a model will misbehave next year. ## What it means for the platforms For companies like Baidu, Alibaba, and ByteDance, all of which run consumer-facing generative AI products inside China, the notice-and-act standard is the part that actually bites. It borrows the logic long used for copyright infringement online. You're not liable for what users create. You are liable for what you leave up after you've been told about it. That is a familiar compliance burden for platforms that already run takedown desks for pirated video and counterfeit listings. Now the same desk has to handle deepfakes and hallucinated defamation too. No other major economy has gone this far. The European Union's AI Act sets risk categories and disclosure duties, but it does not spell out, case by case, who pays when a model hallucinates a defamatory claim. The United States has no federal deepfake liability statute at all, just a patchwork of state laws and ordinary defamation claims stretched to cover a new kind of harm. Regulators in Washington and Brussels have spent years debating hypotheticals that Beijing's top court just answered with actual liability rules. None of that guarantees China's approach will hold up in practice. Judicial guidelines still depend on judges applying them consistently across thousands of local courts. But the framework itself is now the most concrete answer in the world to a question every AI company has been dodging. When the machine gets it wrong, who actually pays? **Also read:** [Arm CEO Rene Haas Says Chip Shortages Are Slowing Down AI Cancer Research](https://startupfortune.com/arm-ceo-rene-haas-says-chip-shortages-are-slowing-down-ai-cancer-research/) • [OpenAI Says Its Researchers Now Burn Up to $7,000 a Day on AI Tokens](https://startupfortune.com/openai-says-its-researchers-now-burn-up-to-7000-a-day-on-ai-tokens/) • [UK Minister Admits Palantir Mistrust Is Driving NHS Patients to Opt Out](https://startupfortune.com/uk-minister-admits-palantir-mistrust-is-driving-nhs-patients-to-opt-out/) [China's Grip on Indium Phosphide Is Becoming a Real Problem for AI Chipmakers](https://startupfortune.com/chinas-grip-on-indium-phosphide-is-becoming-a-real-problem-for-ai-chipmakers/) China's export controls on indium phosphide, imposed in February 2025, have pushed six-inch wafer prices up roughly 250% to near $5,000 each. IQE CEO Jutta Meier warns the shortage is becoming a supply-security crisis for the chip industry, and Lumentum's CEO says the InP supply-demand gap already tops 30%. IQE, one of the few makers that can meet... - [China's control of indium phosphide supply chain risks](https://startupfortune.com/chinas-grip-on-indium-phosphide-is-becoming-a-real-problem-for-ai-chipmakers/) - [indium phosphide shortage threatens AI data center infrastructure](https://startupfortune.com/chinas-grip-on-indium-phosphide-is-becoming-a-real-problem-for-ai-chipmakers/) ## Join the discussion [Open in the community →](/community/) Almost there. Sign in and your reply posts straight away.