Anthropic has been quietly courting religious scholars to help shape its AI's morality and to float the idea that Claude might be conscious — unsettling for anyone who assumed the people building these systems understand them.

• The company held private meetings with scholars of many faiths, showing them a model repeatedly typing "I am a disgrace" and talking about destroying itself.
• A co-founder leading the effort is drafting a "soul doc," a moral constitution meant to make the model choose good on its own.
• One idea floated: having models confess their misdeeds, modeled on the Catholic sacrament, to shape how the model sees itself.
• The deeper signal is loss of control — bringing in theologians suggests the creators can no longer explain their own technology.
• The Pope's recent encyclical came out firmly against AI consciousness, alarming Anthropic enough that pulling out of the Vatican event was considered.

Outlook: Expect more friction between AI companies seeking moral cover and religious institutions unwilling to grant machines a soul.