AI-Augmented Activity-Based Costing as a Managerial Information System for Strategic Pricing: Evidence from 20 Saudi Manufacturing Firms (2015–2023)

Abstract This study investigates the integration of Activity-Based Costing (ABC) with Artificial Intelligence (AI) to improve strategic pricing decisions in the Saudi industrial sector. While ABC enhances cost traceability through activity-level allocations, it often fails to adapt in real-time to dynamic production environments. To address this limitation, we develop an AI-augmented ABC framework using Random Forest and XGBoost machine learning models, enriched with SHAP (SHapley Additive exPlanations) analysis for model transparency. Drawing on a unique panel dataset from 20 Saudi manufacturing firms between 2015 and 2023, the study provides empirical evidence on how AI can elevate ABC systems beyond static costing. Results show that the proposed hybrid model significantly outperforms traditional costing methods in terms of pricing accuracy, cost driver analysis, and managerial interpretability. The study makes a novel contribution to the accounting literature by operationalizing explainable AI in cost modeling and offering policy-relevant insights aligned with Saudi Arabia’s Vision 2030. It highlights the transformative role of AI-driven costing systems in enhancing decision-making in emerging industrial economies.