Autonomous vehicles are predicted to enter the commercial passenger vehicle market soon. Being a new technology, there is significant uncertainty surrounding its performance, adaptations by users, and influence on the wider transport and energy systems. In this uncertainty are several possibilities for autonomous vehicle technology to underdeliver on its promise of safety, efficiency, and flexibility; and, instead, become a net burden on society. The aim of this study is to develop insights into the medium- and long-term impacts of autonomous vehicles adoption on the use-stage greenhouse gas emissions. In the study, a System Dynamics approach was applied to develop a stock-and-flow model that simulates the technological intervention of replacing a conventional vehicle with an autonomous vehicle under scenarios related to vehicle adoption, induced vehicle demand, vehicle fleet turnover, vehicle ownership, vehicle efficiency, and the energy-supply greenhouse gas intensity. The results indicate that the internal-combustion-engine vehicle fuel efficiency and the electricity-supply greenhouse gas intensity have the largest influence on decreasing greenhouse gas emissions, but that the benefits can be negated by inefficient autonomous vehicles and induced demand. Based on these insights, recommendations to manufacturers include minimising energy consumption and adopting low-greenhouse gas vehicle technologies. Recommendations to governments include managing vehicle travel speeds, incentivising ride-sharing, and continuing support of renewable electricity supply. Shared recommendations include co-planning and preparing for data collection and reporting by autonomous vehicles, themselves, to support the facilitation of a low-greenhouse gas transition. The suggested future work focusses on developing the model, possibly in conjunction with other methods and tools, to support further exploration of other greenhouse gas emitting processes. (C) 2020 Elsevier Ltd. All rights reserved.