Abstract

This chapter addresses the challenge of attributing moral responsibility for the actions of AWS. It argues that these challenges can be surmounted if human agents _assume_ meaningful moral responsibility for AWS actions. This responsibility, justified and fair, reflects an individual's moral character. Human agents can choose to design, develop, and deploy AWS despite uncertainties about potential outcomes, accepting the responsibility and blame for any unintended, unwanted, or unforeseen consequences. Given the inherent unpredictability of AWS, human agents who take such a responsibility take a _moral gambit_. The chapter concludes that while a moral gambit is acceptable for non-lethal AWS, it is not permissible for lethal autonomous weapon systems.