This working paper introduces a framework for evaluating trust in molecular information systems by examining the alignment between an observed material state and its recorded provenance. In bio-digital systems such as DNA data storage, synthetic biology, and biofabrication, successful decoding is often treated as evidence of reliability. This paper argues that this is insufficient: a molecular artefact may remain readable while no longer being consistent with its claimed history. The paper proposes state–provenance alignment as a missing evaluative layer for molecular trust. It frames trust as a relationship between the observed molecular state and the expected state conditioned on provenance, expressed through the minimal formulation T(S, P). Through examples including degradation without detection, regeneration mismatch, and metadata–state inconsistency, the paper shows why informational correctness alone cannot establish evidential reliability. The contribution is conceptual rather than a complete computational implementation. It aims to clarify the need for structured reasoning about whether molecular evidence is compatible with the processes, handling conditions, and histories it is claimed to represent.
Raphael Kim (Thu,) studied this question.