In today’s competitive world, precise cost management in projects is vital, and the activity-based costing system has been designed to accurately allocate costs to activities. However, its implementation complexity and the extensive data requirements make its practical application challenging. In the era of digital transformation, the utilization of modern technologies such as machine learning and artificial intelligence in the domains of finance and accounting has become an undeniable necessity. This study, using a qualitative approach and a documentary research method during 2025 with the aid of MAXQDA software, developed a conceptual model to integrate the capabilities of machine learning and artificial intelligence within the framework of activity-based project costing. First, by extracting 20 core components of machine learning, 18 key components of artificial intelligence, and 15 central components of activity-based costing, a conceptual and integrative classification was presented. Then, these components were aggregated and analyzed within conceptual dimensions to identify their systemic and practical relationships in project cost management. The findings of the study indicate that machine learning models can enable precise cost prediction, identification of cost drivers, and analysis of cost behavior; whereas artificial intelligence, through expert systems, fuzzy logic, and intelligent agents, can facilitate decision-making under uncertainty. Ultimately, the integration of these technologies into the activity-based costing structure enhances the accuracy of cost allocation, increases financial information transparency, and optimizes the utilization of resources in projects. Based on the study results, managerial recommendations have been provided for the practical implementation of this model.