How should programmers integrate ‘fairness constraints’ in the programming (training, validation) of AI systems? How should regulators design regulatory frameworks that are conducive to accurate and fair (as well as accurate because fair) AI systems? When harm occurs due to an ‘error’ (typically, a bias), which adjudicatory procedures should be made available to the victims? Which substantive and procedural rights should claimants and defendants be able to exercise? How should courts adjudicate AI-related disputes in order to arrive at fair outcomes?