Mandatory Computational Lineage Tracing & Algorithmic Utility Passport

The advent of highly complex Generative AI models (LLMs, deep neural networks) has shifted critical systemic risk from traditional data breaches and infrastructure failure to 'algorithmic opacity.' This refers to the inability for human operators, regulators, or affected citizens to determine the precise source, weighting, biases, or specific training data provenance that contributed to a high-stakes computational output. As AI drives automated decision vectors—from autonomous resource allocation and credit scoring to medical diagnosis and military targeting parameters—the failure point is no longer merely technical; it is epistemic and legally undefinable when the model's internal logic is opaque. The resulting systemic risk mandates a structural utility layer: the Computational Utility Passport (CUP).