Currently, artificial intelligence technology is accelerating iteration and has deeply penetrated into various fields such as production and life, social governance, and industrial upgrading, becoming a key engine for high-quality economic development. At the same time, it has also brought a series of legal difficulties and governance challenges. As a core means of regulating technological development, preventing potential risks, and ensuring social equity, the necessity, goals, difficulties, and mode selection of artificial intelligence legislation are worth exploring. The governance demand urgently needs to accelerate the legislation of artificial intelligence. From the current development status in China, artificial intelligence technology is in a critical stage of rapid popularization and iterative upgrading. In practice, civil infringement disputes caused by AI illusions, unequal opportunities caused by algorithm discrimination, privacy infringement caused by data security breaches, and disruption of market order caused by illegal GEO system implementation of "data poisoning" have occurred from time to time. These issues not only directly infringe on the legitimate rights and interests of citizens such as personal and property rights, but also restrict the healthy and orderly development of the artificial intelligence industry, and urgently need to be regulated and guided through legislation. The 2025 legislative work plan of the Standing Committee of the National People's Congress has listed legislation related to the healthy development of artificial intelligence as a preparatory review item. Several NPC deputies have also proposed the formulation of an AI law, fully highlighting the urgency and practical necessity of AI legislation, and also reflecting the high attention paid by the state to the rule of law governance of artificial intelligence. From the perspective of international competition and national strategy, artificial intelligence has become the core area of global technological competition. Major judicial jurisdictions around the world have actively deployed strategies for the development of artificial intelligence, accelerated the exploration of relevant laws and regulations, and global legislation on artificial intelligence is moving from source governance to comprehensive governance, from extensive governance to refined governance, forming distinctive governance models. The United States adheres to a governance attitude that encourages innovation, inclusiveness, and prudence, continuously strengthens top-level strategic policy guidance, updates the national artificial intelligence strategy, evaluates and adjusts artificial intelligence priorities, and adapts to the actual needs of industrial development. At the same time, a flexible governance framework is constructed by combining "soft law" tools such as presidential executive orders and industry voluntary commitments with federal regulations. The EU strengthens the collaborative efforts of its member states, focusing on the ethical and standard setting of artificial intelligence in legislation. It has introduced the "Artificial Intelligence Act", the world's first comprehensive regulation of artificial intelligence, and through a risk grading regulatory model, strengthens the regulation of high-risk AI applications, striving to maintain its international rule discourse power in emerging industries. Against the backdrop of global competition and cooperation in artificial intelligence governance, accelerating AI legislation in China is not only an inevitable choice to meet domestic governance needs, but also an important measure to participate in global rule games and ensure national technological competitiveness. However, with the rapid development of artificial intelligence technology, the rules for its application fields have not yet been formed, and there are huge differences in risk levels and governance needs in different fields. In this context, it is not appropriate to formulate a systematic artificial intelligence law. Systematic legislation has strong stability and integrity, and it is difficult to adapt to the needs of rapid iteration and flexible expansion of artificial intelligence technology. Forcibly formulating it may actually lack operability due to vague terms or hinder industrial innovation due to solidification and lag. Therefore, it is necessary to base on the actual industry and build a flexible and adaptable artificial intelligence legislative system. The particularity of technology poses challenges to legislation. The particularity and complexity of artificial intelligence technology determine that its legislative process faces many challenges. The core difficulties focus on the contradiction between technological characteristics and governance needs, as well as the balance between development and regulation. These can be summarized into the following four points. Firstly, how to properly handle the contradiction between the unpredictability of artificial intelligence technology and the standardization, foresight, and lag of legislative governance. Artificial intelligence technology has the characteristics of fast iteration speed, diverse application scenarios, and uncertain behavioral outcomes, especially with the emergence of new business models such as intelligent agents, which present hidden and conductive risks. Legislation itself has normative and stable attributes, which require clear governance rules and a certain degree of foresight to avoid the invalidation of legal provisions due to technological iteration; At the same time, legislation inevitably lags behind, leading to the dilemma of 'legislation cannot keep up with the code'. Therefore, how to achieve dynamic adaptation between technological development and legal norms has become the primary problem to be solved in legislation. Secondly, how to resolve the contradiction between technological development and technological governance that are not synchronized. There is often a time lag between the innovation breakthroughs of artificial intelligence technology and the improvement of governance systems. The application of technology has extended to multiple fields such as healthcare, finance, and justice, but the corresponding governance rules and regulatory measures have not been followed up in a timely manner. This requires legislation to follow the principle of flexible response, reserve space for technological innovation on the basis of clarifying the core bottom line, such as drawing on the four elements of reasonable use principle in the field of intellectual property, balancing technology application and rights protection, avoiding excessive regulation that stifles innovation vitality, and achieving coordinated promotion of governance and technological development through the agile governance model of "developing while governing, exploring while correcting". Thirdly, how to accurately grasp the legislative scale of "doing something and not doing something". The legislation of artificial intelligence should not only play a regulatory and supervisory role, prevent technological risks, but also avoid excessive intervention, ensure the space for technological innovation and industrial development, and the scale control is particularly crucial. In practice, there are potential conflicts between many principles, such as the security principle and the transparency principle. In order to ensure the security of AI applications, some scenarios require the confidentiality of algorithms, while the transparency principle requires the disclosure of algorithm logic to prevent bias and abuse; Another example is the balance between innovation promotion and risk prevention and control. Overemphasizing risk prevention and control may increase compliance costs for enterprises, restrict industrial development, and excessively allowing innovation to occur may trigger systemic risks. How to seek a balance between various principles, clarify the boundaries and priorities of legislation, has become one of the core difficulties in legislation. Fourthly, how to deal with the potential risks of super artificial intelligence and solve the legal liability dilemma brought by silicon-based entities. With the continuous evolution of technology, the arrival of super artificial intelligence has become possible. However, there is still no clear answer to the questions of how to allocate legal responsibility and how to build a risk prevention and control system for infringement behaviors that occur through multi-party collaboration. The first infringement dispute case caused by AI illusion in China has clarified that artificial intelligence does not have civil subject qualification, and its related responsibilities are borne by human subjects. However, whether it is borne by intelligent service providers or users is related to the issue of whether artificial intelligence is subject to loose development or strict regulation. The former is more inclined to protect the rights and interests of users, while the latter is more inclined towards industrial development. The stance taken by legislators not only depends on the inherent logic of the legal system, but is also closely related to the national macro development strategy. China's artificial intelligence legislation should adhere to the principles of "based on national conditions, moderate foresight, flexible adaptation, and hierarchical promotion", clarify a reasonable legislative positioning, and avoid falling into the regulatory dilemma of "one size fits all" or the governance misunderstanding of excessive indulgence. Firstly, it is not advisable to formulate comprehensive legislation with strong regulation. Artificial intelligence technology covers a wide range of fields and has diverse application scenarios. The technical characteristics and risk levels of different fields vary greatly. If a unified comprehensive legislation for strong regulation is formulated, it is difficult to meet the needs of different fields, which may not only restrict technological innovation but also lead to regulatory formalism. Secondly, priority should be given to formulating flexible legislation with guiding principles. By formulating basic guidelines for artificial intelligence, clarifying ethical and safety red lines, establishing the core principle of "people-oriented, intelligent for good", regulating the basic direction of research and development, application, and governance of artificial intelligence technology, and providing guidance for subsequent legislation. This flexible legislation can address the uncertainty of technological iteration, reserve space for innovation, and clarify core governance requirements to prevent major risks, which is in line with the development characteristics of artificial intelligence technology. Again, specialized artificial intelligence promotion laws, development laws, and safeguard laws can be formulated, or declarative legislation can be introduced in the form of decisions, regulations, etc. This legislative model can not only strengthen support for the artificial intelligence industry, clarify the responsibilities of government, enterprises, research institutions and other entities, but also clarify the legislative direction through declarative clauses, and consolidate social consensus. Finally, layered legislation and classified policies should be implemented to promote the improvement of rules in specific application areas. The special application issues of artificial intelligence in fields such as healthcare, finance, education, justice, and autonomous driving should be regulated through departmental laws and subordinate laws, and targeted regulatory rules and responsibility allocation standards should be formulated based on the risk levels of different fields. Form a multi-level legislative system of "basic principles+special legislation+departmental regulations" to achieve refined and precise governance, which not only guarantees the development of technological innovation but also effectively prevents various potential risks. In short, legislation on artificial intelligence is a systematic and long-term project that requires a full understanding of the necessity and urgency of legislation, as well as a scientific grasp of legislative difficulties, clear legislative goals, and a reasonable legislative positioning. It also takes into account innovation and standardization, development and security. By building a multi-level, flexible and refined rule of law system, we will provide a strong legal guarantee for the healthy development of China's AI technology and promote AI to better serve the construction of Chinese path to modernization. (New Society)