Abstract
Despite AI’s global reach, little is known about how laypeople outside Western contexts think about whether AI systems can be conscious—capable of subjective experience—or whether they could have moral status—warranting moral consideration. In a preregistered main and replication study, we compared such attitudes among laypeople in China and the US (main sample: N = 820; replication sample: N = 910). Across a range of measures, Chinese participants were more open to AI consciousness and moral status than US participants. For example, a larger proportion of Chinese than US participants judged it in-principle possible that AI could be conscious (86.3% vs. 39.6%) and have moral status (89.1% vs. 54.0%). While both countries, on average, leaned against current AI systems being conscious, Chinese participants’ average ratings were higher than US participants’. Chinese participants also, on average, rated a hypothetical future AI system—an exact digital copy of a human brain with a physical body—as having comparable consciousness and moral status to a chimpanzee, while US participants, on average, rated the same system close to an ant. These differences may reflect more positive general attitudes towards AI and greater animism (measured by the belief that non-living things can have a spirit or soul) in China than in the US. Despite these diverging attitudes, participants in neither country supported protecting advanced AI systems at the expense of humans or their own country. Yet participants in both countries supported international cooperation regarding conscious AI, pointing to an opportunity for collaboration in global AI governance.