In a move that underscores China’s cautious approach to integrating artificial intelligence into its intellectual property system, the All-China Patent Attorneys Association (ACPAA) has imposed a strict ban on its members using AI to generate patent application documents. Announced on September 22, 2025, this revision to the Code of Professional Ethics and Practice Discipline for Patent Agents aims to elevate professional standards and foster high-quality development in the patent agency sector. The policy explicitly prohibits patent agents from employing generative AI tools to directly produce application materials, reflecting concerns over authenticity and ethical practices.

The ban also extends to curbing “malicious” price competition among agents, which the ACPAA views as detrimental to industry integrity. According to details outlined in a recent report by The National Law Review, this measure builds on earlier regional restrictions, such as Nanjing’s Intellectual Property Protection Center’s prohibition on AI-generated content in patent filings to ensure document authenticity and impose penalties for violations.

Regulatory Context and Precedents

China’s regulatory framework for AI in patents has been evolving rapidly. Earlier this year, the National Intellectual Property Administration (CNIPA) issued guidelines for patent applications involving AI-related inventions, categorizing them into four types and emphasizing eligibility criteria like human inventorship. These guidelines, detailed in a January 2025 update from China IP Law Update, prohibit AI from being listed as inventors, aligning with global norms while promoting innovation.

This ACPAA ban isn’t isolated; it follows Nanjing’s June 2025 directive, which banned generative AI in drafting patent documents for pre-examination, with strict verification processes. Practitioners have criticized such measures as overly broad, potentially hindering efficiency, as noted in a June 2025 analysis by Managing Intellectual Property, where sources argued enforcement could prove challenging.

Implications for Innovation and Global Standing

The restrictions come amid China’s dominance in AI patent filings, with the country holding 62% of global AI patents from 2010 to 2022, far surpassing the U.S. at 21%, per a 2024 report from IP CloseUp. This lead, highlighted in a 2023 Bloomberg article on China’s widening edge over the U.S., stems from Beijing’s tech drive, yet the new bans signal a pivot toward quality over quantity.

Industry insiders worry that prohibiting AI in drafting could slow patent processing, especially as CNIPA explores AI for examination, as mentioned in their June 2025 press conference covered by The National Law Review. Posts on X reflect mixed sentiments, with some users praising China’s regulatory foresight on IP protection in AI training data, echoing 2023 discussions on legal datasets, while others decry it as a barrier to innovation.

Ethical Concerns and Future Directions

At the heart of the ban are ethical issues, including the risk of AI-generated content lacking originality or infringing on existing IP. A March 2025 Chinese court ruling, reported by King & Wood Mallesons, deemed AI-generated images non-copyrightable due to insufficient human authorship, converging with U.S. standards and underscoring broader debates.

For patent attorneys, compliance means relying on human expertise, potentially increasing costs but ensuring accountability. As China balances innovation with regulation—evident in its draft AI policies requiring legal data sources, as noted in 2023 X posts on generative AI—the ACPAA’s move may influence global practices. Experts suggest this could lead to hybrid models where AI assists but doesn’t generate core documents, fostering a more robust IP ecosystem.

Broader Industry Reactions and Challenges

Reactions from the field vary. Some practitioners, as quoted in Managing Intellectual Property, view the ban as unwarranted, arguing AI can enhance drafting without replacing human judgment. Meanwhile, CNIPA’s “Blue Sky” campaign, which rectified 35 firms for abnormal patent applications in June 2025, per The National Law Review, highlights ongoing efforts to curb abuses, including those potentially amplified by AI.

Globally, this contrasts with more permissive approaches in the U.S., where AI patent eligibility is under scrutiny, as explored in a 2024 article from Fish & Richardson. China’s strategy may strengthen its IP regime but risks alienating innovators if enforcement stifles tools like AI for preliminary research.

In the long term, these developments position China as a leader in ethical AI integration for IP, potentially setting precedents for international standards. As the sector adapts, patent agents must navigate these rules while leveraging technology responsibly, ensuring the high-quality development the ACPAA envisions.