Industry: Balancing Incentives for Innovation and Safeguarding Rights – Exploring Intellectual Property Protection in the AI Era

Artificial intelligence (AI) can write novels, create paintings, generate videos, and even "invent" new technologies. Should it be granted intellectual property rights? As technology evolves rapidly, finding a balance between incentivizing innovation and protecting original creators' rights has become a shared challenge for the global industry and institutional designers.

Against the backdrop of intensifying global technological competition and AI's profound reshaping of industrial structures, intellectual property, as a core pillar of innovation-driven development, is seeing its institutional refinement and practical innovation receive increasing attention. Recently, at the 12th Qiangguo Intellectual Property Forum, guests discussing the theme "IP Practice and Institutional Construction in the AI Era" noted that AI is not a replacement for human creativity but an important tool for amplifying innovative value. They emphasized the need to build a new intellectual property ecosystem that effectively protects original works while promoting technological inclusivity through institutional innovation, technological enablement, and global collaboration.

Behind this consensus lies a clear understanding of the nature of AI technology. "AI is not a creative subject; the current system should adhere to 'humanism' and guard against mistaking technological effects for the source of intelligence," said Zhang Qin, Academician of the World Academy of Nuclear Sciences and former Vice Chairman of the China Association for Science and Technology. He stated that generative AI relies on technologies like deep neural networks and Token prediction, and although its operational process may seem transparent, it is essentially a form of "unconscious mathematical fitting."

Zhang Qin believes that AI presents two prominent intellectual property challenges. On one hand, using copyrighted content during the training process may constitute infringement. On the other hand, if AI-generated content lacks substantial human involvement, it should not, in principle, enjoy copyright. Even with limited human intervention, the rights obtained would be minimal.

"Currently, AI and the intellectual property system face four major challenges," said Lyu Guoliang, former Deputy Director of the WIPO China Office. "First is the subject qualification of the protected party, as the current IP system is based on 'natural persons' as rights holders. Second is the blurring of protection standards—how to redefine creativity and originality. Third is the copyright dilemma of data training, where the 'transformative fair use' argued by AI companies is seen by creators as value extraction without compensation. Fourth is the fragmentation of global governance: AI technology is borderless, but IP rules are territorial, potentially exacerbating global innovation inequality."

In facing these challenges, there is a need to actively explore practical and feasible solutions. Yang Xuri, President of the Beijing Qiangguo Intellectual Property Research Institute, suggested enhancing the intellectualization of patent examination and management by applying large language model technology to intelligent classification, semantic search, and infringement identification for patents and trademarks. He also emphasized strengthening infringement monitoring and enforcement, suggesting that platforms need to enhance the deployment of copyright filtering and piracy tracking technologies for AIGC content like films and music. Furthermore, he advised bridging the gap between achievement transformation and industrial application by using AI to match technological supply and demand.

Forum participants expressed that AI presents a profound "pressure test" and an opportunity for upgrading. By adhering to the principle of "human-centricity and technology for good," leveraging intelligent means to enhance governance efficacy, and simultaneously reinforcing the institutional bottom line of original creation protection, we are fully capable of building a new IP ecosystem that both incentivizes originality and accommodates technological evolution. This will not only inject momentum into high-quality development but also contribute China's forward-looking, inclusive, and feasible solutions to global AI governance.