How reactions to AI are shaped by cultural differences
Cultural differences shape how societies perceive and integrate AI into daily life. While individualistic cultures tend to view this technology as a threat to autonomy, collectivist cultures are likely to see it as an extension of the self.
As the debate about the benefits and drawbacks of AI rumbles on, the technology has undeniably become an integral part of society. Its use is now so ingrained that researchers have observed how human behavior is adapting to meet its standards.
But how is the adoption of AI influenced by culturally driven psychological differences in, say, the United States, Spain and China? Do established cultural norms in one country result in a different view of AI than in another? And if so, what impact does this have on the development of the globally influential technology?
Aaron J. Barnes (University of Louisville), Yuanyuan Zhang (Baruch College, City University of New York) and Ana Valenzuela (Esade) have carried out a comprehensive review of existing literature to examine how cultural identity determines responses to AI. Their analysis, published in Current Opinion in Psychology, offers insight into how individuals conceptualize and use AI in decision-making.
The influence of identity
AI’s ability to perform tasks and make decisions on behalf of humans continues to revolutionize consumer relationships. However, its acceptance varies from country to country. In India, Singapore and China, between 50 and 59% of companies have embraced the technology, but only 26-33% of companies in France, Spain and the US have done the same.
In Eastern countries, more people think AI is beneficial for society
The Western trend towards skepticism is also present amongst the public in general—more people in Eastern countries believe AI to be beneficial for society than in Western countries. And while economic factors can have an impact on whether AI is more likely to be accepted, analysis of secondary data suggests that differences in cultural values could play a more significant role.
By viewing the adoption of AI through a cultural psychology perspective and exploring these differences, the researchers have been able to develop a deeper understanding of consumer responses to AI and examine the influence that identity and sense of self have on acceptance.
Social norms
‘Culture’ can take on various meanings depending on the context within which the word is being used. In a geographical sense, the culture of a place suggests a set of shared values and accepted social norms. Individuals within any given society may define themselves by their cultural heritage, and by extension define relationships with people from other backgrounds differently.
Cultural tendencies towards individualism (freedom from collective control) or collectivism (structures that favor groups over individuals) shape people’s views accordingly. The sense of self is reinforced by the culture of the society in which a person lives—and these cultural variations and patterns influence how societies view AI.
Collectivist cultures are more open to accepting external elements as more powerful than themselves
Individualistic cultures are more likely to see AI as external to the self: a piece of equipment that can perform tasks. A Western European who considers themselves to be a good baker, for example, would be distrustful of a bread-making machine and see it as a sub-standard alternative rather than a helpful piece of equipment. But in collectivist cultures more open to accepting external elements as more powerful than themselves — such as those with strong religious beliefs — AI is more likely to be accepted as a beneficial product or service.
Agency over algorithms
The algorithms that power AI usually reduce human behavior and preferences to readable variables that eliminate the unique aspects of a person’s character. In individualistic cultures like the US, where uniqueness and high levels of agency are valued, these overly simplified interpretations of human traits strengthen the aversion to AI. Framing the technology as a way to enhance uniqueness—such as personalized recommendations based on preferences, or paid-for services offering exclusivity—may reduce this effect.
In collectivist cultures where conformity is seen as being the bedrock of a harmonious society—Korea, for example, where the word conformity is synonymous with maturity and strength— algorithmic decision-making is more likely to be viewed as promoting fairness and enhancing social cohesion. This is also the case in countries that have experienced high levels of corruption, where trust in AI algorithms that make decisions based on pre-determined factors is high.
The value placed on the amount of agency a person has over decision-making is also highly influenced by culture. In the US, individuals tend to be driven and internally controlled, leading to an unwillingness to accept decisions made by AI. But in collectivist cultures such as India, where actions are more responsive to external environments, AI is seen as less likely to limit agency.
Full disclosure
The amount of data AI requires in order to function remains one of the most controversial issues surrounding its development and use. As well as gathering huge swathes of information from public sources, personal information is necessary to foster successful consumer relationships.
For collectivists, who lean towards conformity, sharing personal opinions or recommendations is unusual. In individualistic cultures, where self-expression is the norm, divulging personal information is more common. These views directly correspond with opinions on privacy: collectivists are comfortable with data being collected and used to improve the AI offering, whereas individualists who share personal details display greater levels of anxiety over AI’s ability to protect their information.
Overall, the researchers conclude, individualists are more likely to view AI as external to the self, an invasion of privacy and threat to autonomy. The collectivist view is one of an extension of self, acceptance and conformity. However, they acknowledge that these are just two aspects of culturally driven psychological differences and suggest many other opportunities exist to explore the cultural dimensions that influence AI.