Governing artificial intelligence risks for effective public administration: the case of Kuwait

Governance of Artificial Intelligence (AI) is increasingly discussed within public administration, yet risk centred conceptualisations remain more developed than empirical evidence in contexts where digital infrastructure is still lagging. This research examines residents’ perceptions of AI related risks in Kuwait across the four domains of economic, organisational, social and technical, and considers what these patterns imply for governance priorities in public administration. A nationally representative survey of 679 respondents identifies heightened social risks, particularly related to job security and freedom of expression, alongside moderate economic and organisational risks and variable technical risks. Regression analysis indicates that males perceive lower economic but higher social risks, and that respondents aged 41+ perceive lower economic and technical risks. Education has only a marginal effect on social risk perception. The findings underscore the need for stronger, risk informed governance attention, as residents’ evaluations signal where trust pressures and safeguarding expectations are likely to concentrate as digital public services expand. The research concludes by calling for a stakeholder-led governance mission to ensure that innovation is balanced against social welfare in Kuwait.