In the mobility sector, a large number of new technologies such as autonomous vehicles (AVs) and services (e.g. carpooling) are emerging. AVs involve not only passengers, but also authorities, manufacturers, public transportation companies, law enforcement officials, drivers, pedestrians and shopkeepers. Applying phenomenology - the description of a phenomenon's live experience [1] - to this case of Mobility as a Service (MaaS) contributes to understanding its complexity and provides insights of users' perception of risk related to the AVs. This new technology brings many opportunities to improve our mobility system. Identified potential risks can affect the efficiency and the perception of the service. In this exploratory research, we have employed a technique called experimental phenomenology to identify these risks. The major advantage of this approach is to take into account the perception of passengers as a driver for design.