Anthropic's co-founder Christopher Olah hosted religious thinkers at a San Francisco dinner, discussing whether its AI model, Claude, might be conscious. Rabbi Mois Navon noted Anthropic leaders treated Claude as having “moral status” akin to a person. Olah privately warned AI could enable bioweapons in 12 to 18 months.

Anthropic, potentially valued at $2 trillion, held secret meetings with scholars to apply moral wisdom to its models, aiming to ensure AI chooses goodness to protect humanity.