Trust in Moral Machines: How automation, morality, and media framing drive cross-cultural adoption of autonomous vehicles
通过在美国、墨西哥和中国进行大规模实验，研究了自动化水平、道德编程和事故严重性框架如何共同影响人们对自动驾驶汽车的信任，发现文化价值观调节了这些因素的作用。
We conducted a large-scale, tri-national experiment drawing on the Moral Machine clusters—Western individualist (U.S.), Latin-American transitional (Mexico), and East Asian collectivist (China)—to examine how autonomy level (SAE Level 2 vs. Level 5), moral programming (self-protective vs. utilitarian), and accident-severity framing (low vs. high) jointly shape performance trust in autonomous vehicles. In balanced samples of 300 respondents per country, participants were randomly assigned to one of eight conditions and then reported on performance trust, performance risk, hedonic well-being, and behavioral intentions. A robust three-way interaction predicted performance trust in all three contexts, but with opposite patterns: utilitarian framing enhanced the trust advantage of full autonomy under high-severity scenarios in the U.S. and Mexico, whereas in China it did so only when accidents were mild and undermined trust when accidents were severe. These results demonstrate that cultural worldviews condition how ethical programming and risk context interact to shape trust, offering actionable guidance for culturally sensitive AV design and policy. • Three-way interaction of autonomy, morality, and severity jointly altered trust. • Moral framing had culture-specific effects on AV trust under accident scenarios. • US and Mexico see trust drop in severe accidents under self-protective logic. • China sees trust boost in mild accidents under utilitarian logic.