Tech Giants Seek Religious Wisdom for AI Ethics, Experts Urge Caution

As the race to develop advanced artificial intelligence accelerates, technology companies are increasingly turning to religious traditions for guidance on instilling moral principles into their AI models. Executives are grappling with how to teach chatbots right from wrong, seeing religion as a millennia-old source for answers to questions that a laboratory setting cannot fully address.

This engagement with faith communities comes amid widespread concerns about AI's potential for harm. Companies aim to create morally sound AI by drawing from a broad aggregate of moral ideas found across various traditions. OpenAI CEO Sam Altman has suggested that AI should reflect a "weighted average" of humanity's moral views. However, experts and religious adherents caution that this approach may oversimplify complex theological and ethical systems and fail to account for the inherent contradictions between different worldviews.

Scholars like Webb Keane, an anthropologist at the University of Michigan, argue that religion is often viewed too narrowly as a set of extractable rules, divorced from its cultural, political, and practical contexts. He suggests that AI companies might inadvertently perpetuate a simplified understanding of religion, potentially prioritizing certain traditions, particularly Abrahamic faiths, over others or distorting the essence of faiths that do not primarily define themselves as ethical systems.

Despite these concerns, some companies are reportedly engaging with religious leaders and scholars in thoughtful ways. Meghan Sullivan, a philosopher of religion leading an ethics and AI initiative at the University of Notre Dame, noted Anthropic's rigorous approach to consultations, which included diverse Christian denominations. She described meetings where the company explained its AI models and received feedback on moral questions, ranging from the nature of human souls to the appropriateness of AI for educational purposes.

Representatives from Meta, Google, and Amazon have also engaged with the Vatican, and Microsoft has actively pursued religious outreach. These discussions often occur within broader philosophical debates about the nature of religion itself. Scholars point out that the modern concept of distinct, comparable world religions, often seen as universal, is a relatively recent construct, solidified by Western academic traditions in the 19th century. Religions like Buddhism and Hinduism, for instance, were not historically perceived as clearly bounded traditions by their practitioners before the advent of European scholarly categorization.

This historical context is significant, as it informs why Silicon Valley might adopt a broad, generalized approach to religion. The pursuit of diverse viewpoints can lead to religion being used as a proxy for mapping global diversity of thought. However, critics, such as Paolo Carozza, a legal expert and member of Meta's Oversight Board, worry that this aggregation risks homogenizing distinct traditions. He suggests that AI's method of processing vast datasets could lead to a superficial amalgamation of beliefs, potentially losing the nuanced, organic development characteristic of living traditions and favoring Western religious perspectives.

Anthropic has stated its goal is not to align AI with any single tradition but to draw from a wide range of viewpoints—religious, secular, and political—to understand how "good character actually forms." Nonetheless, the challenge remains for tech companies to navigate the complexities of religious ethics without reducing diverse traditions to simplistic moral algorithms.