OpenClaw, open source AI agent Photo: VCG
The China National Intellectual Property Administration (CNIPA) issued a risk alert on Wednesday, warning that AI-powered agent tools such as OpenClaw were found to have weak default security settings, which could lead to serious security risks. CNIPA said that the use of such intelligent AI agents in drafting patent application documents could pose multiple risks, including leakage of technical information, lack of content in patent applications, and loss of credibility for patent applicants.
In a post posted on its official WeChat account, the top IP regulator pointed out that agent AIs such as OpenClaw pose risks such as excessive privileges, security vulnerabilities, and plugin poisoning.
Using these AI tools to create patent application documents can easily lead to the leakage of core information such as technology disclosure documents. Once leaked, the technical solution in the patent application may lose its novelty and become ineligible for approval, or worse, it may be pre-emptively filed by others, causing significant losses to the applicant. In addition, the agency could be held liable for breach of contract, the agency noted.
In addition, when such AI tools are used to prepare patent application documents, problems such as AI illusions, logical contradictions, and unclear explanations of technical features may occur, which may result in the application not receiving protection from IP authorities.
For anyone who uses an intelligent agent to fabricate, randomly generate, or stitch together content for a patent application, that constitutes malicious patent filing activity. If certain criteria are met, applicants may be subject to administrative penalties such as warnings and fines. Pursuant to CNIPA, agents and patent agents may be subject to penalties such as revocation of their business licenses and revocation of their certificates of competency.
In serious cases, they may be included in the list of serious and unreliable entities, CNIPA warned.
To reduce risks and protect the legitimate rights and interests of all parties, CNIPA urges applicants to increase their awareness of risk prevention, carefully select compliant patent agency services, and proactively check whether their agencies use agent AI to draft application documents.
If an applicant discovers that an agency has used such tools without permission, resulting in information leakage or malicious application activity, the applicant may file a complaint and seek compensation in accordance with the law.
Patent offices and patent agents should always be vigilant about the risks associated with the use of intelligent agents, refrain from using intelligent agents to engage in malicious patent application practices, and effectively protect the legitimate rights and interests of their clients, read the post.
As AI agents such as OpenClaw become more popular, many Chinese government agencies are warning of hidden risks.
On March 13, the National Cybersecurity Notification Center issued a risk alert regarding OpenClaw, warning that numerous exposed OpenClaw assets pose significant security risks and are a prime target for cyberattacks.
Global Times
