Governance of Augmented Thought: A Semi-automated BPMN Framework for Verifiable Decision-making and Human-ai Workflows in Higher Education

Rojas Caqui A.J.; Arbildo Prieto D.H.A.; Esaine Ortiz N.M.; Leon Lucano J.J. · Proceedings of 2026 International Conference on AI Decision-Making and Management, AIDMM 2026 (2026) · DOI: 10.1145/3811839.3811863

Resumen

The rapid adoption of Generative AI has introduced challenges in traceability and verification. This study proposes the CT-IA/ATF Framework, a computational architecture based on BPMN 2.0 standards for the optimization of semi-automated data flows. The system implements a multi-lane execution logic that separates human-machine processing units. Central to this architecture are three algorithmic control gates: (G1) Automated Traceability Mapping, (G2) Logic Consistency Verification via cross-reference protocols, and (G3) Incremental Contribution Analytics. Using Design Science Research, we evaluate the system's latency, inter-rater reliability, and the reduction of computational hallucinations through a structured audit-trail database. Results indicate that this decision architecture significantly improves process reliability and reduces algorithmic governance risks.

Instituciones: Universidad Continental; Universidad de Lima