Zhao Jingwu

In the era of digital economy, artificial intelligence technology has become an important technological turning point for the innovative development of the future digital economy, and "artificial intelligence+industry scenarios" has once become the mainstream trend of technological development. Of course, AI technology cannot avoid the double-edged sword effect. In recent years, the Administrative Provisions on the Recommendation of Algorithms for Internet Information Services, the Administrative Provisions on the Deep Synthesis of Internet Information Services, and the Administrative Measures for Generative AI have been promulgated, with the purpose of preventing various security risks caused by technology abuse and promoting the safe and reliable development of AI. In order to promote governance in this field in the future, it is necessary to follow the basic concepts of comprehensive governance and systemic governance, integrate multiple governance tools such as technical standards, markets, laws, and policies, and form an ecological system of scientific and technological innovation. Among them, the role of law in promoting technological innovation cannot be solely attributed to conclusions such as "clarifying the legitimacy boundaries of technological innovation" and "promoting the transformation of technological innovation achievements". Instead, a systematic mechanism for promoting technological innovation should be constructed from three aspects: risk prevention, market incentives, and innovation resource guarantee.

The reason for the transformation of legislative concepts is that the governance concept and mechanism of artificial intelligence should be synchronized with industrial practice and market demand. In the early stages of artificial intelligence development, this type of technology was still being commercialized for the first time, and its special security risks exceeded the scope of current legislation. Therefore, at this stage, the corresponding legislative focus was mainly on security risk prevention. Subsequently, specialized legislation related to artificial intelligence was successively promulgated, and a preliminary safety risk prevention system has been formed. At this time, the focus of legislative work needs to return to achieving another legislative goal, which is to promote the innovative development of artificial intelligence technology. Objectively speaking, the governance rules extended by the legislative purposes of risk prevention and promoting innovation have significant differences. The legislative purpose of promoting innovation emphasizes more on meeting the practical needs of scientific and technological innovation through corresponding policy incentive clauses and innovative resource supply clauses in accordance with market allocation mechanisms. Moreover, this legislative logic differs from specific rules such as the Law on Promoting the Transformation of Scientific and Technological Achievements and the Law on Scientific and Technological Progress, and requires more institutional construction that is tailored to the practical characteristics of the artificial intelligence industry, focusing on risk prevention, market incentives, and innovation resource protection. Therefore, the focus of legislation on artificial intelligence should be changed, and on the basis of existing industrial development needs, the principles of safety and reliability, economic benefits, and infrastructure protection should be clearly defined.

The principle of safety and reliability refers to the fact that the development and application of artificial intelligence technology should meet the requirements of ensuring that technical safety risks are controlled within an acceptable range. One is to comply with safety technical standards and legal safety management obligations. The innovative methods of artificial intelligence technology should comply with the mandatory provisions of laws and regulations on technical security protection; Meanwhile, in specific application scenarios, priority should be given to adopting technical architectures and communication protocols that comply with general security technology standards. The second is that it can effectively prevent common technical security risks and quickly restore basic functions in network security incidents. Thirdly, it meets the requirements of scientific and technological ethical review. The application of artificial intelligence technology should be based on protecting human dignity and freedom, and conduct ethical reviews in various technological innovation stages such as design and development, commercial applications, etc.

The principle of economic benefits refers to the fact that the promotion mechanism of artificial intelligence technology innovation should meet the requirements of market economic benefits. Encouragement and promotion clauses for technological innovation should emphasize the basic requirement of "market regulation as the mainstay and government intervention as the auxiliary", and connect the functional positioning of market adjustment mechanism and government intervention mechanism in promoting the process of artificial intelligence technology innovation. One is to protect technological innovation achievements and their intellectual property rights, which requires clarifying the intellectual property rights of innovation subjects towards artificial intelligence technology innovation achievements at the level of basic legal principles. The second is to coordinate the intellectual property demands of different innovation entities. The legislative content should provide more specific guidelines for the use of training data, such as specific regulatory requirements for generating content, prohibiting the use of "works" contained in training data to generate similar or identical works, etc. The third is to clarify the principle of connection between government intervention and market incentives, that is, government intervention in artificial intelligence technology innovation needs to respond to specific market practice needs such as investment and financing introduction in the research and development stage, supply of technology research and development sites, provision of cooperative innovation platforms, and transformation of scientific and technological achievements.

The principle of infrastructure guarantee refers to the promotion mechanism of artificial intelligence technology innovation, which should include the necessary supply of innovative resources for artificial intelligence technology innovation. Algorithms, computing power, and data have become the three core elements that affect the innovation of artificial intelligence technology, and the corresponding innovation resource supply mechanism should also be constructed around these three elements. At the algorithmic level, it is necessary to encourage and support higher education institutions and research institutions to carry out scientific research on basic algorithm models, as well as to systematically cultivate key talents in the field of basic algorithm models. At the level of computing power, regulatory agencies need to plan and deploy the construction of computing power center nodes uniformly at the national level to avoid situations where the overall computing power resources are limited due to disorderly construction of data centers. On the data level, on the one hand, the legislative content should clarify the basic principles of fair supply of public data, ensuring that the artificial intelligence industry can obtain necessary training data resources under the same conditions; On the other hand, legislative content should also include specific ways of providing public data, such as building central and local AI industry training data supply platforms.

In summary, the role of law in promoting technological innovation should comprehensively construct a systematic mechanism for promoting technological innovation from three aspects: risk prevention, market incentives, and innovation resource guarantee. Technological innovation has the characteristics of complexity and systematicity. In the case of market regulation failure and limited risk prevention, it is also necessary to pay attention to the guarantee of innovative resources. In the face of specific technological innovation needs, institutional norms should be used to ensure the availability and accessibility of innovative resources.

(The author is an associate professor at the School of Law, Beihang University)

Source: Rule of Law Daily, Rule of Law Network