The Impact of Generative Artificial Intelligence on Decision-Making Quality: A Comparative Cross-Country Analysis
Year: 2026
Type: article
Abstract: Generative artificial intelligence (GenAI) is increasingly used for information retrieval, analysis, content generation, and decision support. However, cross-country evidence linking GenAI use directly to decision-making quality remains limited because available secondary datasets do not measure these constructs with a common instrument. This study therefore uses a multi-source secondary-data design and distinguishes direct GenAI measures, broader AI measures, decision-support activities, decision-related outcomes, and country-level responsible-AI indicators. The empirical analysis uses a 900-observation U.S. repeated cross-sectional GenAI survey covering three waves (2024–2026) and the 2026 Global Index on Responsible AI (GIRAI) country-level dataset. In the U.S. survey, GenAI use frequency is positively associated with a decision-support activity index based on getting information, generating ideas or solutions, and asking advice (N = 896; robust OLS coefficient = 0.711, SE = 0.022, p < .001, 95% CI [0.667, 0.754], R² = .547; standardized β = .733). The analysis does not interpret this association as evidence of improved organizational decision quality. GIRAI provides contextual evidence on AI literacy, human oversight, transparency, and skills for seven of the eight countries in the original study scope, while Iraq is not represented in the supplied GIRAI country file. The findings therefore support a narrower conclusion: direct GenAI evidence is available for some decision-support activities, but the supplied secondary evidence does not justify a unified cross-country causal claim that GenAI improves decision-making quality.
Source: Computer and decision making.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7222888024