Anthropic's Olah courted by the Vatican — and he says they're right about AI risks

anthropic cofounder chris olah spoke at a vatican ceremony following pope leo's ai encyclical, stating that every frontier ai lab faces incentives that can conflict with ethical behavior. this collaboration resulted from outreach by santa clara university ethicists brian patrick green and pastor brendan mcguire, who began meeting with olah last fall. mcguire contributed a 28-page commentary based on medieval mysticism to help refine the behavioral parameters in claude's constitution. the engagement highlights the vatican's strategic 'minerva dialogues' to insert moral constraints into the ai industry's agi pursuit.