Automated driving is an emerging transport innovation whose future impact is increasingly studied. With this paper I want to show that such a literature misses a crucial point: which impacts automated driving will generate depends on the competition between different networks of innovators, each supporting its own approach to the integration of automated driving into the system of urban mobility. As technological, business and policy innovations are simultaneously at stake, a socio-technical approach is followed. In particular, socio-technical maps representing urban mobility in year 2040 are used to represent the results of three transition pathways towards automated driving, each being led by a different network of innovators. The first socio-technical map results from the ability of today incumbent automotive companies to integrate the suppliers of technologies and components for connected automated vehicles into their network, and to successfully lobby for public investments on vehicle-to-infrastructure technologies for long-haul mobility. No relevant changes in urban mobility are generated in this transition pathway, except for the diffusion and the peer-to-peer sharing of household electric automated vehicles. The second socio-technical map emerges from the cooperation between leaders of the internet and managers of shared and collective transport systems. This new network of innovators is able to gain support from multilevel policies that consider shared automated vehicles as the core element of diffused systems of integrated urban mobility. Impacts on urban mobility are positive, also because the strict regulation of household automated vehicles avoids any rebound effect. In the third socio-technical map, urban mobility is taken over by energy agents who integrate automated vehicles and smart grids. Such innovators foster a new political discourse on energy efficiency and sustainability, and eventually gain support from the public co-funding of national and international smart grids. In this scenario, unregulated pervasive rebound effects of automated driving may result from the limited attention paid by policies to the issue of urban quality and sustainable transportation.