In many countries, it’s common to see former church buildings turned into apartments, offices, or bars. This trend points to a broad global decline in religious affiliation and identity. The United States has seen one of the biggest drops: the share of its residents who said they belonged to a church, synagogue, or mosque fell from 70 percent in 1999 to 47 percent in 2020, according to Gallup.

One potential factor is the proliferation of artificial intelligence and robotics, according to a team of researchers led by Chicago Booth’s Joshua Conrad Jackson and Northwestern’s Adam Waytz. The more exposed people are to automation technologies, the researchers find, the weaker their religious beliefs. They argue that the relationship is not coincidental and that “there are meaningful properties of automation which encourage religious decline.”

Researchers and philosophers have pondered the connection between science and religion for many years. The German sociologist Max Weber spoke of science contributing to the “disenchantment of the world,” or the replacement of supernatural explanations for the workings of the universe with rational, scientific ones. Evidence from prior research doesn’t support a strong “disenchantment” effect, Jackson says, but he and his coresearchers suggest that AI and robotics may influence people’s beliefs in a way that science more generally does not.

To explore the link between automation and religion, the researchers gathered data at the national, regional, organizational, and individual levels. They combined survey data on the importance of religion with information on the use of industrial robots to compare the populations of 68 countries over the years 2006–19. They controlled for factors such as per capita GDP, population size, and exposure to other technologies. The researchers find that countries with a large stock of robots—1 standard deviation above average—experienced a 3 percent decline in religiosity each decade. By contrast, countries with a stock 1 standard deviation below average experienced a slight uptick in religiosity.

They also looked at members of a single community who self-reported information about their religious beliefs and their jobs at various points over time. Using an occupational database that indicated the importance of programming to various professional roles—a proxy for AI exposure—the researchers find that people whose jobs were 1 standard deviation higher than average in exposure to AI were 45 percent less likely to believe in God than those with average exposure. Further, because they were able to follow responses over time, they observed how a person’s career path and religious feelings changed from survey to survey.

Losing their religion

Countries* where workers had more exposure to robots tended to experience a decline in religiosity, the research finds.

“Occupational AI exposure explained variation in religious belief across individuals, but also religious decline in the same individual over time,” the researchers write.

Two other studies—one comparing religiosity and the stock of industrial robots across US metropolitan areas and the other exploring the relationship between occupational exposure to AI and religious attitudes and practices among employees of a manufacturing plant that was integrating AI technology—both supported the link between automation and religious decline.

In a follow-up experiment, the researchers asked participants to read about either three advances in AI technology or three non-AI scientific developments. The participants then rated how impressive and technologically advanced each discovery was, and also how much it affected the strength of their religious faith. Reading about AI had a stronger negative effect on participants’ religious convictions than reading about other kinds of breakthroughs, even when the breakthroughs were judged as equally impressive.

Automation’s impact on religious faith may have to do with what the researchers call the “instrumental” value of religion. “Historically, people have deferred to supernatural agents and religious professionals to solve instrumental problems beyond the scope of human ability,” they write. “These problems may seem more solvable for people working and living in highly automated spaces.”

Jackson and his coresearchers note that apart from its instrumental value, religion also plays a role in defining morality. Whether people ultimately consider automation a suitable substitute for religion in performing that function is an open question. They point out that people are currently unlikely to trust AI with moral decision-making, but this trust may grow over time.

More from Chicago Booth Review

More from Chicago Booth

Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.