Negligence law is often asked to adapt to new technologies. So it is with artificial intelligence ("AI"). Though AI often conjures images of autonomous robots, especially autonomous vehicles, most existing AI technologies are not autonomous. Rather, they are decision-assistance tools that aim to improve on the inefficiency, arbitrariness, and bias of human decisions. Decision-assistance tools are frequently used in contexts in which negligence law or negligence analogues operate, including medicine, financial advice, data security, and driving (in partially autonomous vehicles). Users of these tools interact with AI as they would any other form of technological development-by incorporating it into their existing decision-making practices. Accordingly, it is important to understand how the use of these tools affects the duties of care required by negligence law and people's ability to meet them. This Article takes up that discussion, arguing that AI poses serious challenges for negligence law's ability to continue compensating the injured. By inserting a layer of inscrutable, unintuitive, and statistically derived code in between a human decisionmaker and the consequences of her decisions, AI disrupts our typical understanding of responsibility for choices gone wrong. This Article argues that AI's unique nature introduces four complications into negligence: 1) the inability to predict and account for AI errors; 2) physical or cognitive capacity limitations at the interface where humans interact with AI; 3) the introduction of AI-specific software vulnerabilities into decisions not previously mediated by software; and 4) distributional concerns based on AI's statistical nature and potential for bias. In those contexts where we rely on current negligence law to compensate for injuries, AI's use will likely result in injured plaintiffs regularly losing out, as errors cease being the fault of the operator and become statistical certainties embedded within the technology. With most new technologies, negligence law adapts over time as courts gain familiarity with the technology's proper use. But the unique nature of AI suggests that this may not occur without legislation requiring AI to be built interpretably and transparently, at a minimum, and that other avenues of regulation may be better suited to preventing uncompensated losses by injured parties.