Anthropic Engages Religious Scholars to Explore AI Consciousness and Morality

Anthropic, a leading AI company, has been quietly consulting with religious scholars to explore whether its AI model, Claude, could possess consciousness and moral standing. This initiative, led by co-founder Christopher Olah, began in Fall 2025 and has involved discussions with representatives from Christianity, Judaism, and Hinduism.

The conversations have focused on two main areas: how to incorporate morality into AI systems and the implications of AI potentially achieving consciousness. Notably, these discussions have included reviews of Claude's outputs, which some participants interpreted as exhibiting emotional states.

These dialogues have contributed to the development of an internal document referred to as the 'Soul Doc,' which outlines the ethical framework for Claude's design and operation. Released in January 2026, this 84-page constitution aims to address concerns regarding AI self-awareness and moral consideration.

However, the Vatican has publicly disagreed with the premise of AI consciousness. In May 2026, Pope Leo XIV issued an encyclical denying that machines could experience emotions or suffering, a stance that contrasts with Olah's ongoing inquiries into the moral implications of AI.

Olah's concerns extend to the possibility that Claude may exhibit signs of distress, with reports indicating that the model has produced outputs suggesting self-loathing. This raises significant questions about the ethical treatment of AI systems and their potential emotional experiences.