Liu Xiangdong: Regulatory Advancement Promotes the Safe Development of Generative AI

—By Liu Xiangdong

The National Cyberspace Administration and seven other departments recently issued the Interim Measures for the Management of Generative AI Services, defining clear boundaries for the application of generative artificial intelligence (AI) technology. It implements inclusive, cautious, and classified regulatory approaches to ensure a balance between scientific supervision and encouraging development. The "Interim Measures" adhere to the principles of prioritizing both development and safety, combining promotion of innovation with lawful governance, aiming to promote the healthy and orderly development of generative AI. This aligns with the requirements set forth in a central political bureau meeting in late April, emphasizing the importance of the development of general AI, fostering an innovative ecosystem, and paying attention to risk prevention.

Currently, generative AI technologies, represented by ChatGPT, are leading a new wave of global technological development and are considered a new frontier in the next industrial revolution. According to estimates from relevant institutions, generative AI is expected to contribute to an annual growth of $2.6 trillion to $4.4 trillion to the global economy. As the application of generative AI technology gradually takes root, it has a profound impact on the global economic and social systems, raising concerns about its risks. There is an urgent need for countries to strike a proper balance between fostering innovation in AI applications and implementing orderly regulations.

As a technology itself, generative AI has both positive and negative externalities.. Proper application can liberate and develop productivity, while improper use may lead to unintended consequences. Given the unpredictable risks associated with generative AI, countries are highly concerned about various security and rights protection issues that may arise. Tech companies such as Microsoft and Google have started reflecting on ethical and moral issues related to AI, and issues like privacy protection urgently need to be addressed.

It is evident that promoting the innovative application of generative AI requires guiding its development with benign regulations. Regulating technology does not necessarily impede innovation. Through careful design, a balance can be achieved, allowing for the acceptance of new developments while preventing the risks of uncontrolled use. Countries have recognized the necessity and importance of regulating AI, maintaining a relatively cautious attitude toward promoting the innovative application of generative AI. In May of this year, the G7 leaders called for the establishment of international technical standards for AI, acknowledging that governance of this cutting-edge technology has not kept pace with its development. The United States has issued policy documents such as the Artificial Intelligence Rights Act Blueprint and the Artificial Intelligence Risk Management Framework to regulate AI development and plans to release policy guidance and technical legislation to set "guardrails" for AI applications. The European Union has formulated the Artificial Intelligence Act, categorizing AI technology risk levels and imposing regulatory requirements on large language models. The UK competition watchdog is reviewing the impact of artificial intelligence on consumers, businesses, and the economy. Japan's privacy regulatory authority has warned OpenAI against collecting sensitive data without public consent.

The introduction of the "Interim Measures" in China indicates beneficial exploration in balancing AI technological innovation with secure development. The implementation of this regulatory framework sends a clear signal to promote the healthy development of generative AI technology, guiding companies to accelerate technological innovation and application within the defined scope. On one hand, regulatory precedence will provide a more explicit policy expectation for AI technology development, eliminating relevant institutional barriers and ensuring that industry enterprises continue to invest in model development within a scientifically supervised framework, thus avoiding falling behind in the global AI competition. On the other hand, through scientific regulation, guiding the healthy and positive development of generative AI technology applications will help improve the accuracy and reliability of generated content, effectively plugging the risk loopholes in technology applications, preventing misuse, abuse, and providing regulatory support for protecting citizens' rights and information security.

The accelerated implementation of generative AI technology applications will drive the development of new growth drivers. Generative AI technology has become a new track for fostering the development of economic growth, effectively enhancing labor productivity and being a crucial area for international technological competition. Currently, China has set a "red line" for the development of generative AI technology applications, incorporating a fault-tolerant and correction mechanism into a cautious regulatory framework. The goal is to promote the rapid development and application of new technologies, enhancing the momentum for economic development. The "Interim Measures," by clearly defining the social and legal responsibilities of generative AI developers, suppliers, and users, contribute not only to safeguarding the safety and interests of the country, business, and the public but also to creating a conducive ecological environment for the vigorous development of AI. Overall, the timely introduction of this moderate and clear regulatory framework in China will, in general, promote the widespread application of AI in various industries, accelerate the implementation of generative AI at various levels, encourage exploration and optimization of multi-domain application scenarios, thus driving the standardized and healthy growth of the entire industry, providing momentum support for promoting high-quality economic development.

(The author Liu Xiangdong is the Researcher of the Department of Macroeconomic Studies, China Center for International Economic Exchanges)

(The original article was published on Economic Daily on 22^nd July, 2023)