Survey study reveals institutional legitimacy drives public acceptance of safety AI in South Korea, indicating governance readiness outweighs pure technical accuracy.
Key Points
To examine how institutional legitimacy and governance conditions influence public acceptance of AI-based threat detection systems beyond technical accuracy.
Survey-based design administered to 510 valid respondents in South Korea.
Evaluated using a two-step structural equation modeling approach (confirmatory factor analysis and structural path analysis) with 5,000 bias-corrected bootstrap mediation resamples.
Trust in government significantly predicted institutional legitimacy (ß = 0.660, p < 0.001), which subsequently shaped performance expectancy and perceived social deterrence.
The full structural model accounted for 84.8% of the variance in behavioral intention to support safety AI systems (R² = 0.848), confirming 11 of 12 direct hypotheses.
Ethical concerns did not directly undermine institutional legitimacy, demonstrating conditional rather than automatic normative resistance from the public.