The accumulation and severity of traffic accidents are influenced by three dimensions: exposure to accidents, risk, and consequence (Nilsson, 2004). It is expected that automated vehicles (AVs) of high technological readiness would have a lower risk of accidents compared to manually driven vehicles(MVs). However, assessing the influence of AVs on changes in exposure and mode shifts is not straightforward, and various AV concepts that will exist in the future may have different effects on mode shifts. For example, if AVs provide travel options for the first or last miles to public transport, they could shift traffic from passenger cars to public transport, and thus further increase safety. Conversely, if only private AVs (automated passenger cars) are available and there is no enhancement in the quality of public transport, travelling by car may increase and shift users away from public transport, which currently presents a very low accident risk for its passengers. Consequently, despite lower accident risk per kilometre driven than for MVs, private Avs might even increase the overall accumulation of road fatalities and injuries due to increased exposure. The aim of this study is to develop a conceptual model that illustrates the impacts of mode shifts on road fatalities and injuries within the entire transport system when private AVs enter the transport system. The model considers nine impact mechanisms (IMs) (Kulmala, 2010) that describe both direct and indirect changes associated with AVs across the three dimensions of safety (Innamaa et al., 2018). The model can be used as a basis for future analysis that also utilizes quantitative data