Pope Leo encyclical urges world to ‘disarm’ AI and curb tech power

Leo XIV’s first encyclical is the most comprehensive moral and political statement on AI we’ve seen from a global institution, and it lands squarely in the middle of an aggressive frontier-model race. By explicitly calling for AI to be “disarmed” — stripped of roles in lethal decision-making and removed from unchecked military and economic power — the document tries to reset the default from ‘move fast and break things’ to ‘move carefully under oversight.’ It names the core structural risk correctly: extreme concentration of data, compute, and decision authority in a tiny group of AI firms and political actors.([apnews.com](https://apnews.com/article/d92d0108730d146baa46da041b8523da))

For the race to AGI, the encyclical doesn’t change model capabilities, but it does change the politics around them. It offers language and framing that regulators, civil-society groups, and even cautious industry leaders can borrow when pushing for binding guardrails on lethal autonomy, high‑risk deployments, and corporate power. Think of it as soft power that could harden into law: if the “disarm AI” frame sticks, it will be harder for governments to justify unconstrained military or surveillance use of frontier systems. At the same time, the document explicitly calls for international cooperation rather than moratoria, which means it is more about redirecting the race than stopping it.