AI ADOPTION IN MALAYSIA'S FINANCIAL SERVICES SECTOR: EXTENDING THE TECHNOLOGY ACCEPTANCE MODEL WITH COST AND SYSTEM QUALITY DIMENSIONS

Abstract Malaysian financial service providers—particularly those operating outside the mainstream banking sector—have been slow to integrate Artificial Intelligence into their core operations, despite compelling evidence of its transformative potential in financial management, credit assessment, and risk control. To illuminate the specific forces that shape this reluctance, the present study develops and tests an augmented version of the Technology Acceptance Model that incorporates two underexplored determinants: System Quality and operational Costs. Survey data gathered from 384 microfinance operators spanning Malaysia's northern, central, and southern regions were subjected to Partial Least Squares Structural Equation Modeling. Every postulated relationship achieved statistical significance. Cost proved to be the dominant driver of adoption intention (β = 0.402, p < 0.001), trailed by Perceived Usefulness (β = 0.253, p < 0.001), System Quality (β = 0.056, p < 0.001), and Perceived Ease of Use (β = 0.034, p < 0.001). Together these predictors account for 78.4 percent of variance in adoption intention—an exceptional yield by IS research standards. The ascendancy of Cost over traditionally dominant constructs represents a theoretically noteworthy departure from patterns documented in technologically mature markets, pointing instead toward the structural economic constraints unique to small financial businesses in an emerging market setting. The paper closes with concrete guidance for FinTech solution providers, regulatory authorities, and financial institution managers seeking to narrow Malaysia's AI adoption gap.