An enhanced governance measure for deep synthesis applications: Addressing the moderating effect of moral sensitivity through message framing

Abstract

The risk of malicious deep synthesis lurks in hedonic applications, yet people tend to be ethically tolerant, leaving governance in a quandary. This study explores the reason for this from the perspective of moral judgment and finds that hedonic applications reduce people's sensitivity to ethical norms. We then explore the mechanism of the governance measure labels, using moral sensitivity as a moderating variable, and find that people with low moral sensitivity are immune. Finally, we propose and validate the inclusion of message framing as an enhancement. This study explains the governance dilemma from a moral-psychological perspective and provides practical governance improvements.