Despite its futuristic premise, the debate surrounding the potential punishment of artificial intelligence (AI) systems has achieved remarkable intellectual depth and complexity over the years. In this paper, I aim to taxonomize the competing views on “AI punishment” through the metaphor of a courtroom trial in which an AI system stands as the defendant. Drawing on a real-world example, I imagine a conversational AI system that allegedly drove a user to take his own life being “accused” of “incitement to suicide.” Into this setting enter the prosecution and the defense, embodying two opposing strands of the literature, both grappling with (a) the criminal capacity question (is AI the right sort of entity to punish?), (b) the criminal liability question (can AI systems fulfill the actus reus and mens rea doctrinal requirements of criminal liability?), and (c) the punishment question (even if conceptually feasible, is it desirable to punish AI systems?). The short—and dominant—answer is currently “no.” Yet the debate is worth examining because it compels the interested parties in the trial to reveal their implicit philosophical assumptions as to the overarching purpose of criminal law and its concept of the responsible “person.”