The purpose of the article is clarifying the concepts of de-recognization, anonymization, pseudonymization, and de-identification with the development path on making regulations, and promoting the dynamic equilibrium between the protection of personal information and the full use of data. Since iterative algorithms may re-recognize seriously incomplete anonymous data sets, it is objectively difficult to achieve an absolute state that cannot be recognized and cannot be recovered. It is necessary to ensure that completely anonymous data does not belong to the category of personal information, and conditionally exempt the processor from obtaining consent obligation under fully meeting the basic standards of de-identification. Through the governance mechanism and the protection mechanism of personal information de-identification based on risk controlling and management, and the ownership mechanism of rights and interests on personal information de-identification with data utilization as the core, rights and interests of specific natural persons will be protected, and the value of massive data will be fully realized. These will promote the harmonious development of economy and society.