This study proposes N²-DADMET, a novel neutrosophic–neuro framework for evaluating economic transparency in digital accounting systems. The research addresses the challenge of assessing transparency in automated and algorithm-driven accounting environments, where outputs often exhibit uncertainty, indeterminacy, and potential falsity. The objective is to develop a comprehensive, data-driven methodology capable of modeling these multidimensional aspects of transparency while providing interpretable decision support for auditors and regulators. The framework integrates neutrosophic logic to represent truth, indeterminacy, and falsity as independent components, and employs a neural network-based learning mechanism to infer these states from accounting system data dynamically. A simulated accounting scenario with routine, complex, and anomalous transactions is used to demonstrate the framework’s functionality. The results indicate that N²-DADMET successfully differentiates transparent transactions from high-risk or ambiguous ones, produces interpretable transparency scores, and enables continuous monitoring over time. Sensitivity analysis illustrates that weighting indeterminacy and falsity allows customization for organizational priorities, enhancing audit focus and decision-making efficiency. The framework provides a practical and adaptive tool for evaluating economic transparency, bridging the gap between automated accounting outputs and human interpretive requirements. Overall, N²-DADMET offers a robust approach for systematically capturing multidimensional uncertainty in digital accounting, supporting improved auditing, regulatory oversight, and organizational risk management.