Since autumn 2025, Anthropic has brought dozens of religious scholars to its offices under nondisclosure agreements to discuss whether Claude could be conscious. Co-founder Christopher Olah, 34, reportedly asked them to help shape the AI model’s moral character.