The increasing deployment of autonomous vehicles (AVs) necessitates a clearer understanding of responsibility distribution in accidents involving these technologies. This paper examines the fatal Uber AV accident in Tempe, Arizona, through the lens of the SAE-J3016 standard and the Ontology of Autonomous Driving. Using formal knowledge representation, we analyze the roles of key actors to assess liability in this case. Our findings reveal that the Uber AV failed to detect and appropriately react to a pedestrian crossing the road, while the human operator was assigned a role that did not align with their actual capabilities within the system’s operational constraints. Despite broader systemic failures, only the operator faced criminal charges, which created a discrepancy between legal interpretations and technical reality. By leveraging the ontology, we demonstrate how structured conceptual frameworks can clarify liability in AV incidents. Our analysis underscores the importance of integrating ontological tools into argumentation in legal contexts and regulatory frameworks to ensure a more precise and just assessment of responsibility.