Algorithm Interpretation Right—The First Step to Algorithmic Governance

Abstract

As artificial intelligence technologies are increasingly deployed in this digital era entitled to generate automated and semi-automated decisions, and the internal logic of machine learning algorithms is typically opaque, the absence of a right to explanation will put the individuals in a weak position, so the right to an explanation for such decisions has become a critical legal issue. The right to an explanation is first written in the "General Data Protection Regulations", and this year, China lays out the "Personal Information Protection Law", which also includes the right to explanation. However, there are still shortcomings in the relevant regulations. Therefore, this paper comprehensively sorts out the right to explanation by studying the relevant literature and using the method of comparison, points that to construct the applicable logic of the right to explanation in China to ensure its effective implementation, it is necessary to clarify the connotation of the right to explanation, establish its application path and build a collaborative governance system.

Untitled

“…These tools are at a very early stage of development, but there are already early approaches to predicting court decisions. For this, for an open case, as many comparable historical cases as possible are identified from the database and an estimate is made based on the relative frequency of success cases (Zou & Zhang, 2022). However, decision prediction has still not been ground-breaking, and it is also questionable whether it will ever be possible in specific case categories.…”

Legal Tech: Unravelling the nature and purpose of modern law in the digital era

“…These tools are at a very early stage of development, but there are already early approaches to predicting court decisions. For this, for an open case, as many comparable historical cases as possible are identified from the database and an estimate is made based on the relative frequency of success cases (Zou & Zhang, 2022). However, decision prediction has still not been ground-breaking, and it is also questionable whether it will ever be possible in specific case categories.…”