Existing research on whether people prefer to adopt advice from humans or artificial intelligence (AI) has yielded conflicting findings: Some studies demonstrate people’s preference for human advice, manifesting algorithm aversion, while others show preference for AI advice, indicating algorithm appreciation. Integrating the dual-attitude model and dual-system theory, constructs a cognitive mechanism model, which propose that people hold relatively positive explicit attitudes yet negative implicit attitudes toward AI (vs. human advisors). In addition, intuitive decision-making processes activate negative implicit attitudes towards AI, leading to aversion towards AI advice. Conversely, rational decision-making processes suppress negative implicit attitudes towards AI and activate positive explicit attitudes, resulting in a preference for AI advice. This model systematically explains why and when people prefer AI or human advice, and can integrate contradictory findings from the perspectives of dual attitudes and distinct decision processes. Future research should investigate the dynamic changes in dual attitudes toward AI and focus on implicit attitudes toward AI, thereby better facilitating positive human-AI interactions.

advice taking')">"> advice taking, artificial intelligence, artificial intelligence’s advice