CNIPA Issues Risk Alert on the Use of OpenClaw and Similar AI Agents in Drafting Patent Applications
Published 2 April 2026
Xia Yu
On 1 April 2026, the China National Intellectual Property Administration (CNIPA) published an official notice on its public WeChat account regarding the risks associated with using AI agents such as OpenClaw to draft patent application documents. The purpose of this notice is to respond to growing concerns over the security vulnerabilities and compliance risks of such tools, to raise awareness among applicants and patent practitioners, and to safeguard the integrity of the patent system as well as the legitimate rights and interests of all parties involved.
Risk of Technical Information Leakage
AI agents like OpenClaw may suffer from weak default security configurations, excessive system permissions, and vulnerabilities such as malicious plugins or data exfiltration risks. When these tools are used in drafting patent applications, they often process highly sensitive technical disclosures, including invention details that have not yet been made public.
If such information is leaked, the consequences can be severe. The invention may lose its novelty and therefore become unpatentable. In addition, third parties may exploit leaked information to file competing applications in advance, causing irreparable harm to the original applicant. In such cases, patent agencies may also face contractual liability and compensation claims due to failure in safeguarding confidential information.
Risk of Substantive Defects in Patent Applications
The use of AI agents in drafting patent documents may introduce so-called “AI hallucinations”, where the generated content appears plausible but is factually incorrect or internally inconsistent. This can lead to logical contradictions within the application, unclear or ambiguous definitions of technical features, and even inclusion of non-existent or inaccurate technical solutions.
Such defects directly undermine the quality of the patent application and may result in rejection during examination. Even if granted, patents with unclear or flawed claims may be difficult to enforce, thereby weakening their commercial value and legal protection.
Risk of Dishonest or Non-Compliant Filings
Another significant concern highlighted by CNIPA is the risk of dishonest patent filings. When AI tools are used to generate content without a real technical basis, such as fabricated inventions, randomly assembled descriptions, or artificially constructed solutions, such filings may violate the principle of good faith under patent law.
Engaging in such practices can lead to serious administrative consequences. Applicants may face warnings and fines, while patent agencies and practitioners may be subject to revocation of licenses or qualifications. In severe cases, parties involved may be included in official records of serious legal violations and dishonesty, which can have long-term reputational and operational impacts.
Recommendations for Risk Prevention
CNIPA emphasizes that applicants should strengthen their awareness of these risks and exercise caution when selecting patent service providers. It is advisable for applicants to proactively confirm whether AI tools are being used in the drafting process and ensure that proper safeguards are in place. If unauthorized use of such tools leads to information leakage or constitutes dishonest filing behavior, applicants have the right to file complaints with relevant authorities and seek compensation for damages.
For patent agencies and patent attorneys, the notice underscores the importance of maintaining strict professional and ethical standards. Practitioners are urged to remain vigilant about the risks associated with AI tools, avoid any use that could result in non-compliant or dishonest filings, and take full responsibility for protecting client confidentiality and ensuring the quality and legality of patent applications.
Comment
From the perspective of an intellectual property lawyer, this notice reflects a timely and necessary regulatory response to the rapid integration of AI tools into patent practice. It makes clear that while AI can enhance efficiency, it also introduces new layers of legal and operational risk that cannot be ignored. Patent drafting is fundamentally a high-stakes, detail-intensive process that relies not only on technical accuracy but also on strict confidentiality and legal judgment. These are areas where unchecked reliance on AI tools may create significant vulnerabilities.
The notice also reinforces the principle that responsibility ultimately rests with human actors. Neither applicants nor patent professionals can shift liability to AI systems when issues arise, whether in the form of data breaches, poor-quality drafting, or dishonest filings. This clarification is particularly important in setting expectations for accountability in an era of increasing automation. For practitioners and applicants alike, the message is clear: AI should be used as a supporting tool under careful supervision, not as a substitute for professional expertise and legal responsibility.
Risk of Technical Information Leakage
AI agents like OpenClaw may suffer from weak default security configurations, excessive system permissions, and vulnerabilities such as malicious plugins or data exfiltration risks. When these tools are used in drafting patent applications, they often process highly sensitive technical disclosures, including invention details that have not yet been made public.
If such information is leaked, the consequences can be severe. The invention may lose its novelty and therefore become unpatentable. In addition, third parties may exploit leaked information to file competing applications in advance, causing irreparable harm to the original applicant. In such cases, patent agencies may also face contractual liability and compensation claims due to failure in safeguarding confidential information.
Risk of Substantive Defects in Patent Applications
The use of AI agents in drafting patent documents may introduce so-called “AI hallucinations”, where the generated content appears plausible but is factually incorrect or internally inconsistent. This can lead to logical contradictions within the application, unclear or ambiguous definitions of technical features, and even inclusion of non-existent or inaccurate technical solutions.
Such defects directly undermine the quality of the patent application and may result in rejection during examination. Even if granted, patents with unclear or flawed claims may be difficult to enforce, thereby weakening their commercial value and legal protection.
Risk of Dishonest or Non-Compliant Filings
Another significant concern highlighted by CNIPA is the risk of dishonest patent filings. When AI tools are used to generate content without a real technical basis, such as fabricated inventions, randomly assembled descriptions, or artificially constructed solutions, such filings may violate the principle of good faith under patent law.
Engaging in such practices can lead to serious administrative consequences. Applicants may face warnings and fines, while patent agencies and practitioners may be subject to revocation of licenses or qualifications. In severe cases, parties involved may be included in official records of serious legal violations and dishonesty, which can have long-term reputational and operational impacts.
Recommendations for Risk Prevention
CNIPA emphasizes that applicants should strengthen their awareness of these risks and exercise caution when selecting patent service providers. It is advisable for applicants to proactively confirm whether AI tools are being used in the drafting process and ensure that proper safeguards are in place. If unauthorized use of such tools leads to information leakage or constitutes dishonest filing behavior, applicants have the right to file complaints with relevant authorities and seek compensation for damages.
For patent agencies and patent attorneys, the notice underscores the importance of maintaining strict professional and ethical standards. Practitioners are urged to remain vigilant about the risks associated with AI tools, avoid any use that could result in non-compliant or dishonest filings, and take full responsibility for protecting client confidentiality and ensuring the quality and legality of patent applications.
Comment
From the perspective of an intellectual property lawyer, this notice reflects a timely and necessary regulatory response to the rapid integration of AI tools into patent practice. It makes clear that while AI can enhance efficiency, it also introduces new layers of legal and operational risk that cannot be ignored. Patent drafting is fundamentally a high-stakes, detail-intensive process that relies not only on technical accuracy but also on strict confidentiality and legal judgment. These are areas where unchecked reliance on AI tools may create significant vulnerabilities.
The notice also reinforces the principle that responsibility ultimately rests with human actors. Neither applicants nor patent professionals can shift liability to AI systems when issues arise, whether in the form of data breaches, poor-quality drafting, or dishonest filings. This clarification is particularly important in setting expectations for accountability in an era of increasing automation. For practitioners and applicants alike, the message is clear: AI should be used as a supporting tool under careful supervision, not as a substitute for professional expertise and legal responsibility.