The past decade has seen a systemic shift in the automotive landscape and the constituent parts of a vehicle. The automotive industry has shifted from a primarily hardware components industry to a software heavy industry, with software controlling majority of the vehicle functions. Coupled with the ability to fully update or evolve a vehicle's capabilities or functionalities, post point of sale through software updates, the technical, commercial and service landscape of the automotive industry is rapidly changing. This has brought increasing focus to the concept of Software Defined Vehicle, where the vehicle is not only constantly evolving, but is also becoming more personalised by leveraging data collected through the life of the vehicle. This requires a rethink of the current development and deployment approaches for vehicles, which are software-intensive. In this paper, we introduce a novel four-step system engineering framework for the safe development and deployment of Software Defined Vehicles. Inspired by the System-Theoretic Accident Model and Processes (STAMP) & System-Theoretic Process Analysis (STPA) methodology, we analyse hardware, software, organisational (OEM and suppliers) and governmental factors involved in the development and deployment of SDV as an exhaustive socio-technical system analysis. Taking a first principles approach, our novel approach combines system engineering and controls engineering concepts. Our framework involves: 1) step 1: identification of stakeholders (hardware components, software components, organisational etc.), 2) creation of a nested control structure for identified stakeholders, 3) step 3: listing of assumptions, and step 4) monitoring of assumptions through life cycle of the vehicle. Our framework provides analysis of for the pre-sale (development) phase and post-sale (deployment) phase of the SDV. Furthermore, we present three case studies for three use cases of Software-Defined Vehicles (including ADAS and automated driving), identifying requirements for various stakeholders and illustrating the practical implementation of our novel framework.