Pope Leo XIV warns on AI power in 'Magnifica Humanitas'

With “Magnifica Humanitas,” Pope Leo XIV just elevated AI governance from a niche policy topic to a front-page moral question for 1.4 billion Catholics and a much wider global audience. The encyclical explicitly links AI to questions of human dignity, labor, inequality and war, calling for AI to be “disarmed” from logics of domination and for robust legal and institutional guardrails. That kind of language will echo in parliaments, boardrooms and courtrooms for years, shaping how political coalitions talk about risk, liability and acceptable use.

Strategically, the document creates an unusual alliance between parts of the AI safety community and a major religious institution. By putting Anthropic’s Chris Olah on stage at the rollout, the Vatican is effectively blessing a safety-first posture that often clashes with both Big Tech growth narratives and national security hawks. This complicates the Trump administration’s aggressive pro-military AI stance and gives cover to regulators who want to slow or redirect the most dangerous deployments, especially in autonomy and surveillance.

For the race to AGI, the encyclical doesn’t change who has the most compute or the best models, but it does sharpen the legitimacy battle. Labs that can frame themselves as aligned with this “human-first” ethic may find it easier to secure public trust, while those pushing unconstrained agentic systems could face stronger backlash and, eventually, harder law.