Model transparency: Core principles in AI system design

A panel will lead a dialogue with all members in attendance on a draft of a Commentary which will explore what it means to have an “explainable” AI system, leveraging existing guidance provided by global regulators and industry groups, and will set forth foundational design principles that should be incorporated into legal standards governing AI transparency. The Commentary will articulate what the law should require regarding what, how, and to whom to disclose as it pertains to AI-based systems that rely on personal information to make decisions.