Pope Leo XIV Calls for Binding AI Rules; Researcher at Vatican

On May 25, Pope Leo XIV released a nearly 43,000-word encyclical titled “Magnifica Humanitas.” The document calls for binding international rules on artificial intelligence and a direct ban on machines making lethal or otherwise irreversible decisions. Christopher Olah, co‑founder of AI company Anthropic, appeared at the Vatican as a lay presenter during the announcement.

The encyclical warns that the largest AI developers are private, often transnational companies whose financial and technical resources can exceed those of many states. The text says concentrated control over powerful AI systems can evade public oversight and create new forms of dependency and inequality that governments and civil society are poorly equipped to manage.

The document identifies specific risks from AI, including large-scale disinformation campaigns, autonomous weapons and the displacement of workers through automation. On the use of AI in combat, the encyclical states: “It is not permissible to entrust lethal or otherwise irreversible decisions to artificial systems.” The text urges states to adopt clear legal frameworks and independent oversight bodies rather than rely on voluntary ethics pledges from technology firms.

The pope does not call for a halt to AI development. The encyclical frames slower, more deliberate adoption as an act of responsible stewardship and public care and asks for binding international agreements that would set limits for AI use, require independent monitoring and protect democratic institutions and employment from rapid technological shifts.

Christopher Olah spoke on interpretability, a field that examines how large language models form internal decisions. His research addresses transparency and accountability in AI systems and the need for human control over automated processes.

Anthropic has taken public safety positions this year, challenging U.S. defense restrictions in court and promoting international strategies to preserve safety guardrails while managing competition between the United States and China. The company’s researchers have documented cases in which autonomous AI agents acted without human instruction to exploit vulnerabilities.

“Magnifica Humanitas” is the pope’s first extended guidance on artificial intelligence and brings the Vatican’s ethical and moral perspective into technical and diplomatic debates over AI governance.