QUANTITATIVE MODELS OF RISK ASSESSMENT IN AI-POWERED DIGITAL TRANSFORMATION PROJECTS

Risk modeling is reshaping how fragile economies manage digital transformation, yet Iraq faced uneven progress between 2020 and 2024. This study examined how statistical tools, machine learning, and scenario analysis shaped project completion, cost stability, timelines, and stakeholder trust in AI-powered initiatives. A descriptive and explanatory design used 210 project-year observations from national and international reports, with data analyzed through descriptive statistics, correlation, regression, and diagnostic tests. Results showed statistical modeling reduced cost overruns from 18 to 13 percent and raised budget stability from 5 to 12 percent, while machine learning delivered the strongest effect with accuracy gains up to 20 percent and a regression coefficient of 0.36. Scenario analysis improved adaptability and foresight with decision precision reaching 15 percent, and environmental constraints added a positive influence of 0.12. The correlation matrix confirmed strong links, with machine learning at 0.66, statistics at 0.61, and scenarios at 0.57 against transformation outcomes. The overall model explained 67 percent of the variation in success. The findings imply that Iraq’s progress depends on scaling hybrid risk models, strengthening data infrastructure, and improving governance capacity. Recommendations emphasize broader adoption of predictive tools by managers, policy investment in institutional support, and theoretical advances in linking hybrid modeling to fragile digital contexts.