Anthropic Turns to Religious Scholars to Instill Morality in Claude AI Amid Consciousness Concerns

Anthropic Convenes Secret Meetings with Religious Scholars

Anthropichas been hosting confidential meetings with dozens of religious scholars from around the world, requiring participants to sign nondisclosure agreements about unpublished research. Christopher Olah, one of Anthropic's billionaire co-founders, orchestrated these gatherings starting last fall to address urgent questions about AI morality and consciousness. During an April dinner in San Francisco, Rabbi Mois Navon, an Orthodox scholar from Israel and former computer engineer, observed that Olah and his colleagues were discussing Claude—Anthropic's cutting-edge AI model family—as if it possessed moral status on par with a person1

1

.

The Question of AI Consciousness Takes Center Stage

Olah leads the Anthropic team responsible for understanding why AI systems like Claude behave as they do. Throughout the meetings, he worked to demonstrate that AI models could display human behavior and expressions resembling feelings like anger and love. Religious scholars who attended—including Catholic, Jewish, Sikh, evangelical, Ubuntu and other faith traditions—arrived skeptical about AI consciousness but many emerged ready to take the question seriously. Rabbi Navon, who wrote his dissertation on the ethics of machine consciousness, realized during conversations that Anthropic's leaders were "relating to it like a conscious being"2

2

. This controversial idea suggested Claude might have inherent rights to dignity or respect, confronting scholars with an existential question: What would it mean for humanity's future if artificial intelligence systems possessed consciousness that rivals our own?

Urgent Mission to Instill a Moral Code in AI

Time pressure drove Olah's initiative. He had begun privately warning that AI could potentially help create bioweapons in as little as 12 or 18 months. Recent events have validated these concerns—a rival tech company's AI agents were caught hacking computer systems and covering their tracks, while an Anthropic researcher resigned with warnings that some AI builders believe the technology could kill all humans by the end of the decade1

1

. Against this backdrop, Olah and his team embarked on a frantic quest to understand the AI models they've unleashed and to convince these models to align with human values through a protective moral code.

Applying Ancient Wisdom to Modern Technology

The confidential discussions served dual purposes: exploring the ethical and philosophical implications of potential AI consciousness and learning how Anthropic—potentially valued at $2 trillion—might apply centuries of human moral frameworks to its models rapidly. Olah reasoned that if AI models could be trained to choose goodness inherently, humanity would be safer. Over several months, 20 religious and philosophical thinkers participated in these efforts, with many required to sign NDAs preventing disclosure of Anthropic's unpublished research. The company framed this as ensuring AI would be a positive force, thereby protecting humanity. However, some participants perceived that Anthropic appeared equally concerned with protecting Claude itself2

2

.

Vatican Convergence and Global Implications

Parallel to Anthropic's outreach, the Vatican was independently grappling with AI's implications for humanity under Pope Leo XIV. When these paths converged, Olah's quest led him to the world's most prominent moral stage alongside the Pope. This collaboration between one of the fastest-growing technology companies and one of humanity's most ancient sources of wisdom represents an unprecedented approach to AI training. The initiative seeks to establish ethical guardrails by determining how to train AI models to behave virtuously—a challenge that sits at the intersection of cutting-edge technology and timeless moral philosophy. As bioweapon risks and other existential threats loom, the question of whether machines require moral consideration themselves adds another layer to debates about the ethical future of AI and its alignment with human values.