About 15 religious scholars and ethicists met with AI company Anthropic in late March to discuss how to help the Claude chatbot behave “properly.”
The goal is not to turn Claude into a religious chatbot, but to leverage thousands of years of ethical thinking from faith traditions to shape AI behavior.
Experts argue that AI’s power is outgrowing the internal ethical controls of tech companies, forcing them to seek external support.
On May 25, Pope Leo XIV released a 42,300-word encyclical titled “Magnifica humanitas,” calling for the “disarmament” of AI rather than granting default governance rights to technology.
Anthropic is developing a “Claude constitution,” a set of principles that helps the AI self-inspect and correct its responses according to ethical standards instead of just adhering to a rigid list of rules.
The values the chatbot is trained on could affect hundreds of millions of users on sensitive topics such as abortion, end-of-life care, and overcoming emotional loss.
Some scholars warn that AI companies could use religion as an “ethics washing” tool to create a moral image without necessarily changing actual behavior.
Anthropic later expanded the discussions to various other faiths, including Judaism, Hinduism, Mormonism, Sikhism, and Greek Orthodoxy, to diversify ethical perspectives.
📌 Conclusion: The debate shows that AI is no longer just a technical issue but is becoming a dilemma of morality and social power. Anthropic wants to use religious philosophy to build a “character” for Claude, while the Vatican warns that technology should not automatically have the right to rule over humans. However, AI companies turning to religion also raises concerns about the manipulation of faith and the lack of transparent oversight mechanisms.
