We commend the AI HLEG for developing an Assessment List of this kind. In doing so, the AI HLEG has taken on the extremely difficult task of trying to operationalise a theoretical, ethical framework into a useable format for AI developers and deployers. We believe strongly that this type of Assessment List can help establish real, ethical practices for all AI actors. Though the Assessment List is “primarily addressed to developers and deployers of AI systems,” we also hope that consumers (e.g., government agencies) and end-users of AI systems will use the list to help inform their own thinking about the types of questions that they will want to have answers to before establishing how much to trust particular AI systems. In the attached comments, we provide two high-level comments for the consideration of the HLEG and Commission. In addition, we have also provided numerous concrete edits and suggestions to the list itself in the second portion of this document.

eu-hlegfeedbackhaggstromtegmark.pdf

Professor Max Tegmark  and Professor Olle Häggström