Publication Details
Abstract
The application of Generative Artificial Intelligence (Gen-AI) in enterprise process mining has presented new prospects of adaptive decision-making, optimization of efficiency and preemptive governance. The high pace of the implementation of AI-guided automation poses significant questions associated with transparency, accountability, and ethical handling of intelligent systems. The study introduces a Governance and Accountability Framework of Generative-AI-Assisted Process Mining with the focus on human control, algorithmic accountability, and flexibility of operations within industrial processes. Based on practical operational data of a flotation plant contained in the Quality Prediction in a Mining Process dataset, this paper illustrates how AI-generated insights can be used to develop a better understanding of the process, predict quality deviations, and act as an aid to corrective interventions. The multivariate temporal organization of the dataset allows modeling dynamic work process behaviors and creating simulated workflow scenarios. A tiered governance model is created and implemented with these elements; model interpretability, audit traceability, ethical compliance verification and human-in-the-loop validation. The suggested framework will make the AI-assisted process recommendations answer to the organizational goals and regulatory requirements and will be operationally flexible. The achievements of the experiments point to the fact that the mining workflows supported by Gen-AI are more effective in terms of predicting impurities and minimizing response time, which confirms the integrity of the data-driven decisions. The results of this highlight the idea that an ethical approach to AI design by governance can stabilize automation and ethical control to offer a long-term model of adaptive enterprise systems. This study explores a contribution to the emerging literature on responsible AI governance by filling the gap between process intelligence and generative model responsibility in industrial contexts in actual situations.