The rise of autonomous systems necessitates efficient liability apportionment. Existing approaches, such as Shapley values, are non-causal and computationally expensive. We propose \emph{k-leg liability}, a scalable framework based on Structural Causal Models (SCMs) and robustness semantics of logical specifications, ensuring fair and efficient liability allocation. Our method leverages \emph{causal non-interaction} to decompose computations, reducing complexity. We formally relate k-leg liability to harm quantification and empirically show it closely approximates Shapley values while significantly lowering computational cost. This provides a practical and legally interpretable approach for liability in autonomous systems.