Legal Construction of Algorithm Interpretation

Path of Algorithm Accountability

Abstract

Nowadays the development of AI technology is not yet mature, let alone the legal definition and regulation of its type, even the type of technology itself is full of uncertain factors. Because of the rapid development of technology and the openness of theories, scientists have not yet formed a unified consensus and system on cutting-edge technical issues. Therefore, at present, governments all over the world are actively formulating the development plans of AI, but the supervision and regulation of AI are scattered and lagging behind. There is nothing wrong with encouraging the development of new technologies, but the application of technologies requires a responsible response to various ethical demands from human society. No matter what form of AI technology and its application are inseparable from the algorithm and the issue of “algorithm accountability” may probably be a focus of legal regulations on AI and the path of accountability is algorithm interpretation. It is desirable but regrettable that the EU’s GDPR stipulates the non-binding “right to explanation”. But the stop of GDPR is exactly the starting point of constructing the algorithm interpretation mechanism in law.