The impact of artificial intelligence and predictive analytics on insurance risk assessment in the digital age

Abstract

This cross-sectional study examines the impact of artificial intelligence (AI) and predictive analytics on insurance risk assessment across 10 countries: Ukraine, Kazakhstan, Bosnia and Herzegovina, Poland, Czech Republic, Georgia, Serbia, Uzbekistan, Romania, and Turkey. Utilizing a mixed-methods approach, including a survey of 320 experts and econometric modeling, the research evaluates how AI adoption, predictive analytics usage, and digital infrastructure influence risk assessment accuracy. Results reveal significant regional disparities, with high AI adoption and robust digital infrastructure (e.g., Poland, Turkey) correlating with 74–78% error reduction, compared to 54–58% in lagging regions (e.g., Bosnia and Herzegovina). Regression analysis highlights AI adoption’s positive impact (β = 9.2, p < 0.01), moderated by digital infrastructure (β = 1.12, p < 0.01), while stringent regulations unexpectedly hindered progress (β = -3.1, p < 0.05). Qualitative themes underscore algorithmic bias and infrastructure gaps as critical challenges, particularly in data-scarce contexts. The study produces academic value through its extension of TOE and Diffusion of Innovations frameworks to investigate neglected markets while selecting spatial and infrastructural elements. The investigation provides applications for insurers which include infrastructure capital commitments along with localized methods for reducing bias and increased employee qualifications. Public-private partnerships and adaptive regulations should become policy priorities because they aid the pursuit of balance between innovative progress and fairness in the digital insurance industry.