Abstract
The article examines artificial intelligence as a tool for modernizing public administration and as a potential factor in the economic efficiency of digital transformation. The study aims to develop and substantiate a conceptual-analytical framework identifying channels through which gender-sensitive AI in Kazakhstan’s public sector may be associated with transaction costs, budget targeting, labor-force participation, and returns to human capital, and to specify indicators for subsequent empirical assessment. The methodology combines institutional economic analysis, a comparative approach, policy analysis, elements of gender impact assessment, and logical modelling. The empirical and documentary basis includes official Kazakhstan data on e-government and the labor market, the Law “On Artificial Intelligence,” the Concept for the Development of Artificial Intelligence for 2024–2029, Bureau of National Statistics data, ILOSTAT, ITU DataHub, Women, Business and the Law, and OECD, UNESCO, and UN Women publications. The analysis identifies potential economic channels rather than statistically established causal effects. The scientific novelty is an economic-institutional extension of Algorithmic Accountability, Algorithmic Fairness, Responsible AI, and gender-responsive AI through the concept of gender-algorithmic accountability, which links gender-differentiated AI outcomes with transaction costs, budget targeting, employment, human capital, and returns on public digital investment. Practical implications concern gender auditing of high-risk AI systems, gender- and region-disaggregated monitoring of digital services, fairness requirements in public procurement, and support for women’s economic participation.