关键词: 法律算法化, 算法信任, 数字人权, 可信人工智能

Abstract: Although legal algorithmization is closely related to the concept of trustworthy artificial intelligence (AI), it cannot achieve the credibility of algorithms, and it needs to be combined with digital human rights to jointly solve the problem of algorithmic trust. The existing criticism does not recognize the rationality of legal algorithmization and the inherent quality of law’s pursuit of stability, but only questions the theoretical significance of legal algorithmization from the perspective of legal value. Legal algorithmization realizes that external supervision alone cannot reduce the risk of algorithms, and attempts to control algorithm activities by integrating the law into algorithms to achieve the credibility of algorithms. However, legal algorithmization mistakenly believes that algorithmic trust dilemma only stems from the technical risks of algorithms, and ignores that the emergence of algorithmic trust dilemma is closely related to opacity and imbalance of benefit risk distribution. The immateriality of algorithms also deepens the trust crisis of algorithms. From the perspective of overcoming the trust crisis in algorithms, legal algorithmization should keep enough restraint, develop more legal risk prevention technologies, and less interfere in the distribution of rights and obligations. With the cooperation of digital human rights, legal algorithmization can be turned into an opportunity to develop the cooperative governance model of algorithms, in which legal algorithmization should be optimized.

Key words: legal algorithmization, algorithm trust, digital human rights, trustworthy AI