Abstract: Algorithm application brings about such problems as algorithm discrimination, algorithm collusion and legitimacy of data acquisition, which cannot be effectively solved by post-supervision alone. To eradicate algorithmic problems, the risk prevention and control mechanism has to be improved, and the construction of enterprise data rights system is the top priority. The data possessed by an enterprise has the nature of property due to its huge economic benefits and legal independence due to the value consideration of risk prevention and control; therefore, the enterprise data should be the object of civil rights. The enterprise data rights should be institutionally designed from the three dimensions of data right acquisition, data processing power, and licensed use of data. In the dimension of data right acquisition, it is required to pay attention to maintaining the right of individuals to know and supervise the use of data, so as to solve the issue of the legitimacy of data collection; in the dimension of data processing power, it is required to strengthen the objectivity of data processing and the design of supervision mechanism, so as to regulate algorithm discrimination; in the dimension of licensed use of data, it is required to prohibit the licensed use of algorithm in pricing, so as to solve the problem of algorithmic collusion.

Key words: algorithm discrimination, algorithm collusion, risk prevention and control, the right of data