Scholars question whether averaging religious views preserves each tradition

Some AI developers are asking philosophers and faith leaders how chatbots should handle moral questions as they build more powerful systems. Anthropic co-founder Chris Olah made the case earlier this year at the Vatican’s presentation of Pope Leo XIV’s encyclical on AI: “These are not questions a lab can answer,” he said. “They are questions traditions like yours have carried for millennia.”

The Associated Press reported little public scrutiny of what technology companies think religion is and what they hope to gain from it. It also reported that executives were warning about AI’s destructive potential while debating the need to develop morally good models and increasingly seeking input from philosophers and faith leaders. The outlet described the executives’ philosophy as one in which religion can help inform AI to be morally good amid founded and unfounded fears about the technology.

Anthropic said in a May blog post that the company had met with “scholars, clergy, philosophers and ethicists from more than 15 religious and cross-cultural groups,” a pool it hoped to broaden. The post did not enumerate the traditions represented. “This work isn’t about aligning our models with any one tradition’s worldview. We want Claude to draw from a full range of viewpoints — religious, secular, political — with equal depth and rigor,” Anthropic said. It described the goal as “careful, accumulated thinking on how good character actually forms.”

Because these companies are not bound to a particular religion, the goal is often a broad aggregate of moral ideas that can inform their models, the Associated Press said. OpenAI CEO Sam Altman offered a broader formulation last year. “What I think ChatGPT should do is reflect that weighted average, or whatever, of humanity’s moral view,” he said. Experts and adherents alike have warned that this idea fails to reckon with contradictions among traditions and does not capture how religious belief is lived.

For decades, scholars have pushed back against a widespread view of religion as a set of rules or principles that can be divorced from the “messy, practical, ad hoc” ways adherents live, Webb Keane, an anthropologist at the University of Michigan, argued. “There’s a very conventional way of thinking of religion as a rule book,” he said. “It is certainly very tied to scriptural religion, and the scriptural religions at the end of the day really turn out to be the Abrahamic religions.” Keane warned that assuming religion can offer distinct, extractable moral principles and using it to train AI risks prioritizing certain traditions and distorting others, particularly faiths that do not see themselves primarily as systems of ethics.

The Associated Press reported that the idea religion can be separated from culture and politics remains entrenched even though it suffuses both in many ways. The concern extends to the consultation process itself: asking faith leaders to answer ethical questions may further perpetuate a theory of religion that fails to recognize the importance of doctrine, ritual and relationships.

Although religion is generally considered ancient and universal, experts such as the late Jonathan Z. Smith conceded that it is true in some ways, while many scholars today question whether it can be divided into distinct, comparable systems. Smith wrote in 1982: “Man has had his entire history in which to imagine deities and modes of interaction with them. But man, more precisely Western man, has had only the last few centuries in which to imagine religion.” Many scholars of religion today share Smith’s view of their field, calling the idea a modern notion traced to as late as the 19th century. Tomoko Masuzawa, a religious studies scholar who wrote “The Invention of World Religions,” said Buddhism and Hinduism were not recognized as clearly bounded traditions by the people who practiced them before European scholars arrived. “All these categories were not native,” she said. Western academics solidified those categories for years, only beginning to seriously reexamine them in the latter half of the 20th century.

Those entrenched ideas about religion — that it is easily comparable across cultures — help explain why Silicon Valley might not feel the need to define its own understanding of it, the Associated Press reported. The Associated Press also reported that, in an effort to represent various viewpoints, religion is often an obvious way to showcase diversity of thought. Masuzawa said: “Somehow, religion becomes a way of kind of mapping out the world.” She added: “It’s as if the whole world has kind of bought into it.”

The Associated Press said the meetings had not lacked diverse representation and described the engagement as thoughtful from the perspective of attendees such as Meghan Sullivan, a philosopher of religion who leads an initiative on ethics and AI at the University of Notre Dame. Sullivan attended an Anthropic meeting earlier this year that included Christians from different denominations. Sullivan said she was more impressed by Anthropic’s rigorous engagement with religion than by conversations she had had with other technology companies. “The culture inside the company is similar to that on a good college campus in the sense that you’ve got a collection of very intellectually curious people,” she said. “These meetings are pretty nerdy — I don’t know if that’s the best way to put it — in a good way.”

At Sullivan’s meeting, Anthropic first explained its models “in layman’s terms,” then faith leaders offered feedback and joined discussions. Sullivan described the range: “Every kind of moral question that this technology raises — everything from, ‘Do humans have souls?’ to ‘Should third graders be able to access a chatbot for literacy education?’”

The outreach has extended beyond Anthropic. Representatives from Meta, Google and Amazon have also met with the Vatican, while Microsoft has been vigorous about its religious outreach.

Some people want technology companies to further define their terms and intentions so the public can understand what they expect to gain from religious input. Large language models work by repeatedly guessing the next word in a sentence after having “learned” the patterns of a huge trove of human-written works, including religious texts. Paolo Carozza, a member of the independent Meta Oversight Board and a legal expert who studies Catholic social thought and technology, worried that in the quest for accumulation, a given tradition’s defining characteristics could be lost — in effect, homogenized into something that pulls toward the middle or even favors Western religions.

“Any living, sophisticated tradition over time is capable of interacting with others outside of it and absorbing some new knowledge and growing. But it does so in this organic way,” Carozza said. “What AI, I think, often does and potentially does, is not in an organic way, but in a way that just throws it all together, mixes it all up, and therefore makes sense of none of it.”

Anthropic says its goal is careful moral thinking across religious, secular and political perspectives rather than alignment with one tradition. Keane warns that extracting principles may privilege some traditions; Carozza says accumulating viewpoints may pull them toward the middle.

Associated Press said its religion coverage receives support through a collaboration with The Conversation US, with funding from Lilly Endowment Inc., and that AP was solely responsible for the content.