A reported lawsuit in Zhejiang is putting an unusual form of AI reliance before a local court: a family’s use of a chatbot to choose an auspicious burial date. The case should not be treated as proof that a chatbot caused a later tragedy, or as a ruling that an AI provider is liable. It is a pending dispute built on reported claims, conflicting views of user responsibility, and a public record that does not yet include a judgment.
The South China Morning Post reported on September 20 that a man surnamed Shi filed a claim against an AI company after he used a chatbot to select a burial date for his mother. SCMP attributes the underlying personal account to Jimu News. A report carried by The Paper identifies the service as Doubao (豆包) and provides additional details drawn from local reporting and case information. Neither report is a court decision, a public complaint, or a verified transcript of the conversation.
The lawsuit turns on a reported inconsistency, not established causation
According to Chinese reports, Shi’s family initially selected April 20 with a feng shui practitioner after his mother died in April. Shi later consulted the chatbot, which reportedly recommended April 19. When he asked about the time, the reports say the chatbot suggested a 7 a.m. to 9 a.m. interval while also saying that April 19 was not an auspicious day.
That reported inconsistency forms the basic complaint. It does not establish whether the chatbot’s response was accurately transcribed, what full context preceded the exchange, whether the service made a factual error, or whether the user acted solely on its output. It also does not validate the premise that an auspicious date has any factual effect on later events. The story should preserve the claim as part of the plaintiff’s account, not as a causal mechanism.
Reports say that, after the funeral, a relative was seriously injured in a traffic accident and that some family members linked the event to the burial date. That connection is a reported belief, not an established causal finding. No reviewed source ties the accident to the chatbot, the date, or the funeral. Presenting it as a consequence of AI advice would be both inaccurate and unfair to the parties.
The relief being sought is also limited in the public reporting. Shi reportedly requested an apology and compensation, but no compensation amount has been disclosed. Chinese media reports say the matter was heard by Jiashan County People’s Court on September 3 and was characterized, based on Tianyancha case information, as a network-infringement-liability dispute. The case was reported as ongoing. The reviewed source record contained no official docket, court opinion, complaint, or judgment.
That uncertainty makes this a useful companion to EastFrontier’s earlier coverage of Beijing’s draft rules for interactive AI chatbot services. The present dispute is not a new regulation. It is an unadjudicated example of how users can bring personal expectations and informal decision-making into contact with a generative system.
Current platform terms do not settle a case about earlier events
Doubao’s current user agreement identifies the operator in Chinese as 北京春田知韵科技有限公司. The agreement was updated on September 10 and became effective on September 17. It says AI-generated outputs are for reference, should not be treated as professional advice, and should not be the basis for further action or inaction.
Those terms are relevant to the platform’s current public warning language. They do not establish the terms Shi saw in April, whether he accepted a particular version, or how a court would interpret any clause. The current agreement also does not resolve whether the reported chatbot exchange happened as described, whether a warning was sufficiently prominent, or whether a court would find fault in the software service.
Chinese reporting describes competing positions in the dispute. The company’s side reportedly argued that no online infringement had occurred and that there was no fault in the software service. Shi’s side reportedly challenged the user agreement as a standard-form clause with inadequate warnings and insufficient review of information sources. These are reported litigation positions, not direct quotations from filings or company statements. Neither should be presented as a judicial finding.
The episode illustrates a recurring problem with generative systems. A response can be read as casual information, a suggestion, or an instruction depending on the user’s purpose and belief. In this case, the social importance of a funeral decision amplified the significance the family reportedly assigned to the output. The legal question, however, is not whether an almanac-derived recommendation can cause misfortune. It is how the court evaluates a platform’s duty, the clarity of its warnings, the user’s reliance, and the link between the output and claimed harm.
China’s existing AI rules provide a baseline, not a verdict
China already has a general regulatory framework for public generative-AI services. The Cyberspace Administration of China’s 2023 measure governing public generative-AI services covers systems that create text, images, audio, video, or other content for the public. The measure took effect on August 15, 2023, years before this reported court hearing.
Article 4 requires providers to take effective measures to improve transparency and the accuracy and reliability of generated content. Article 9 requires service agreements with registered users, while Article 10 requires disclosure of appropriate users, settings, and uses, along with guidance for rational use. These provisions describe a regulatory baseline. They do not determine whether Doubao complied with the measures in this case, whether regulators investigated the company, or whether a court will find civil liability.
That distinction matters as China’s discussion of AI-agent accountability evolves. EastFrontier has followed how Hangzhou’s autonomous-governance debate is changing the AI-agent regulation conversation. This case should not be inflated into a landmark precedent or evidence that Chinese law has already solved liability for chatbot output. It is a reported local dispute where the decisive documents remain unavailable.
A separate legal commentary by King & Wood Mallesons describes a different 2025 Hangzhou Internet Court case involving inaccurate university-admissions information, where the provider was reported to have been found not liable after fulfilling a reasonable duty of care. That discussion is not an official judgment and cannot predict the result in Jiashan. Different claims, facts, platform terms, and evidence can produce different outcomes.
The Doubao dispute nevertheless points to an important practical issue. User disclaimers may help explain the intended role of AI output, but they cannot substitute for a factual record of what was said, what was displayed, how it was understood, and what allegedly followed. The source record supports none of the claims that would resolve those questions. Until a court issues an authoritative decision, the responsible conclusion is modest: a reported lawsuit has opened a local test of reliance on chatbot advice, while liability and causation remain unresolved.