Allocation of liability for harm caused at least partially by AI or medical robot can be based upon a binary distinction. The binary is the distinction between substitutive and complementary automation. When AI and robotics substitutes for a physician, strict liability is more appropriate than standard negligence doctrine. When the same technology merely assists a professional, a less stringent standard is appropriate. Such standards will help ensure that the deployment of advanced medical technologies is accomplished in a way that complements extant professionals’ skills, while promoting patient safety.

The 21st Century Cures Act preserved the FDA’s authority to regulate several categories of software that incorporate artificial intelligence/machine learning (AI/ML) techniques. The agency’s draft guidance on Clinical Decision Support Software proposed an approach for regulating CDS software and shed light on the FDA’s approach to genomic software. This chapter explains why the FDA’s proposed approach to regulating software is unlikely to preempt failure-to-warn suits. It then turns to design-defect suits, focusing on adaptive AI/ML software that continues to redesign itself throughout the product lifecycle. How will common defenses work when the state of the art is in the robotic “mind” of an AI/ML algorithm? We also reflect on broader impacts on patients, clinicians, software developers, and healthcare systems and ask whether the FDA is the “right” regulator to address the unresolved ethical and medical practice concerns that surround this software.

If states begin to impose such contractual bargains for automated administrative determinations, the ‘immoveable object’ of inalienable due process rights will clash with the ‘irresistible force’ of legal automation and libertarian conceptions of contractual ‘freedom.’ This chapter explains why legal values must cabin (and often trump) efforts to ‘fast track’ cases via statistical methods, machine learning (ML), or artificial intelligence. Part I explains how due process rights, while flexible, should include four core features in all but the most trivial or routine cases: the ability to explain one’s case, a judgment by a human decisionmaker, an explanation for that judgment, and an ability to appeal. Part II demonstrates why legal automation threatens those rights. Part III critiques potential bargains for legal automation, and concludes that the courts should not accept them. Vulnerable and marginalized persons should not be induced to give up basic human rights, even if some capacious and abstract versions of utilitarianism project they would be ‘better off’ by doing so.

International law presumes some common bonds of belief and empathy among all persons. At the level of individual states, which do more for (and to) their citizens than the present international order, even more is required to maintain social harmony and cohesion. A common understanding of a thicker set of constitutional and social commitments is necessary. But what happens when that common understanding breaks down? For example, what if leaders of one party (and their followers) start to consider other parties illegitimate – inherently disqualified from power? Are there certain “tipping points,” or points of no return, when a pluralistic order passes an event horizon toward a black hole of authoritarianism – an unrecoverable decline and collapse of democracy?