Artificial intelligence has penetrated into current social life, not only expressing human wisdom in the form of machines and penetrating into all sectors of human beings, but also bringing many risks and potential crises. The traditional legal system is difficult to achieve effective constraints on this scientific and technological product. At present, AI must be standardized by the way of “particularity”. Based on the dilemma of legal norms and the particularity of AI, the idea of particularity refers to the division of all AI activities and states into fields, and the “particularity” legal constraints on the states of intelligence or objects in different fields of AI or the identities of persons or objects. With the development of AI technology and the improvement of AI intelligence, the path of “de-particularity” corresponding to “particularity” is needed, so that AI can be integrated into the current legal system. Thus, the legal era proposition of legally restricting AI under the current stable legal system can be realized.