Development of AI-Driven Business Intelligence Frameworks for Financial Risk and Liquidity Monitoring in Real Time

Abstract

This quantitative study developed and statistically validated an AI-driven business intelligence framework for real-time financial risk and liquidity monitoring. A non-experimental, explanatory cross-sectional design was employed, and data were collected from 440 finance, treasury, risk, accounting, auditing, business intelligence, information technology, compliance, and data-analytics professionals working across financial and non-financial organizations. A structured five-point questionnaire measured AI capability, BI system quality, real-time data architecture, data governance, technological readiness, human analytical competence, information timeliness, monitoring quality, financial risk detection, liquidity performance, managerial decision speed, and financial resilience. Cronbach’s alpha coefficients ranged from .84 to .92, while composite reliability values ranged from .85 to .93. The measurement model demonstrated satisfactory fit, with CFI = .953, TLI = .948, RMSEA = .038, and SRMR = .041. Financial risk-detection effectiveness recorded the highest mean score (M = 4.14, SD = 0.55), followed by BI system quality (M = 4.12, SD = 0.58) and real-time monitoring quality (M = 4.09, SD = 0.57). AI capability correlated strongly with financial risk detection (r = .69), while real-time monitoring correlated with risk detection (r = .70) and liquidity performance (r = .68). AI capability was the strongest regression predictor of risk detection (β = .34, p < .001), whereas BI quality was the strongest predictor of liquidity performance (β = .25, p < .001). The structural framework explained 64% of the variance in risk detection, 59% in liquidity performance, 49% in decision speed, and 61% in financial resilience. Real-time monitoring significantly mediated the effects of AI capability on risk detection and liquidity performance. XGBoost achieved the strongest predictive performance, recording 90.2% accuracy and an area under the curve of .946. The findings validated an integrated framework connecting intelligent analytics, information timeliness, continuous monitoring, liquidity management, decision responsiveness, and organizational financial resilience.