On March 31, 2026, the China National Intellectual Property Administration (CNIPA) issued a warning against the use of AI agents such as OpenClaw in the preparation of patent application documents (关用OpenClaw 等 Intelligent 体缰書专利 Application请文品的风险 Proposal). Additionally, any unauthorized use by the Patent Office must be reported. CNIPA states that the use of agents can lead to leakage of technical information, serious flaws including AI hallucinations, and unauthorized (malicious) applications.
The translation of the warning is: The original text can be viewed here (Chinese only).
Recently, it has been revealed that intelligent agent tools such as OpenClaw (formerly known as Clawdbot and Moltbot) have weak default security configurations that can easily lead to serious security risks. At the same time, using such intelligent agents to prepare patent applications may also induce multiple risks. Here are some things to keep in mind:
1. Risk of “technical information leak”: Intelligent agents such as OpenClaw have risks such as excessive privileges, security vulnerabilities, and the possibility of malicious plugins. If used to create application documents, it is likely to lead to leakage of important information such as technology disclosure documents. Once leaked, the technical solution in the patent application may not be approved due to loss of novelty or may be patented preemptively by others, resulting in significant losses to the applicant. The agent is also liable for breach of contract.
2. Risk of “material defects”: If such intelligent agents are used to prepare application documents, “AI illusions” may occur, causing problems such as logical inconsistencies in the content of the application documents and unclear explanations of technical features, which may prevent protection.
3. Risk of “fraudulent application”: Creating a patent application by creating an artificial intelligence agent from scratch, fabricating it at random, splicing together content, etc. is a fraudulent patent application that violates the principle of good faith. If a certain number of applications are received, applicants will be subject to administrative penalties such as warnings and fines. Agents and distributors will be subject to administrative sanctions, including revocation of their business licenses and revocation of their agency qualification certificates, and in serious cases, they will be added to the list of entities that have committed serious misconduct and illegal acts.
In order to prevent the aforementioned risks and protect the legitimate rights and interests of all parties, the following precautions are provided.
1. Applicant: Applicants should increase their risk awareness, carefully select compliant patent agency services, and proactively understand and confirm whether their patent agency uses intelligent agents to draft application documents. If it is discovered that an agency has used related tools without permission, resulting in information leakage or fraudulent application activity, applicants can sue the agency and demand compensation for losses based on the law.
2. Patent Office and Agent: They should be very cautious about the risks of using intelligent agents, refrain from using intelligent agents to file fraudulent patent applications, and effectively protect the legitimate rights and interests of their clients.
