Pope Leo XIV Calls for Binding AI Regulation, Warns Against Profit-Driven Development and Autonomous Lethal Decisions

On Monday, at a special presentation at the Vatican, Pope Leo XIV said that artificial intelligence "now demands to be disarmed," freed from logics that make it an instrument of domination, exclusion and death. He was introducing "Magnifica Humanitas" (Magnificent Humanity), his first encyclical, one of the most authoritative kinds of teaching document a pope can issue. In it he calls for robust regulation of AI and for its developers to work for the common good rather than for profit.

The document has been awaited since the days right after his election, when history's first U.S.-born pope said he considered AI the biggest challenge facing humanity. The text is methodical. A math major himself, Leo traces the history of the church's social teaching and applies its core ideas, justice, solidarity, the dignity of work and the universal destination of resources, to the digital revolution. He signed it May 15, the 135th anniversary of "Rerum Novarum," the key document of his namesake and hero, Leo XIII, which dealt with workers' rights and the limits of capitalism during the Industrial Revolution. The current pope believes AI raises the same existential questions, and "Magnifica Humanitas" becomes the latest in a century-long line of papal updates to that text.

On war, the encyclical is at its strongest. Leo denounces the "culture of power" driving the AI race, especially the push toward ever more sophisticated remote warfare. He says it is "not permissible" to entrust irreversible, lethal decisions to AI systems, and he demands transparency and accountability from developers so that the chain of command behind any AI-assisted strike is always known. He argues that AI has helped accelerate the "normalization of war" by desensitizing people to its cost. He names no specific conflicts, but cites "opposing imperialisms." He also declares the church's just war theory, which sets criteria for when force can be justified, "outdated" given how warfare has changed. That sets up another point of friction with the Trump administration, which has worked aggressively to deregulate AI development.

The encyclical also turns on the companies themselves. Leo repeatedly condemns the concentration of power and data in a few private hands, a danger he says falls especially hard on children and the most vulnerable. "It is not enough to invoke ethics in the abstract," he wrote, calling for robust legal frameworks, independent oversight, informed users and a political system that does not abdicate its responsibility. "A more moral AI is not enough if that morality is determined by a few." On work, he wrote that the pursuit of greater profit cannot justify choices that systematically sacrifice jobs, because the human person is an end, not a means. He appealed to developers and to the political leaders who regulate them to slow down and reflect.

The launch carried an irony. It included remarks by Christopher Olah, co-founder of Anthropic, which is in a legal fight with the Trump administration. The administration ordered all U.S. agencies to stop using Anthropic's technology after the company refused to give the military unrestricted use of it, and Anthropic is suing. The Vatican said it involved the company as part of a decade-long effort to engage Silicon Valley on the human cost of AI. Anthropic and OpenAI are the third- and second-most valuable U.S. private companies, each worth hundreds of billions of dollars, and both are heading toward near-trillion-dollar IPOs.

Olah welcomed the criticism. He said outside checks are fundamental to the technology going well, given "a real possibility that AI will displace human labor at a very large scale." He said more of the world, including religious communities, civil society, scholars and governments, should do what the pope has done, and added: "We need moral voices that the incentives cannot bend." Some critics saw Anthropic's inclusion as a papal stamp of approval. Brian Boyd of the Future of Life Institute read it differently, as akin to a papal audience with a head of state, a recognition of prominence rather than an endorsement. He called Anthropic an enormous corporation taking on enormous risk and responsibility, and said it has shown goodwill, integrity and interest in dialogue. Vatican officials declined to say who contributed to the text.

Experts expect the document to become a benchmark. Paolo Carozza of Notre Dame Law School, who chairs the Meta Oversight Board, called it likely "a defining document for our era." Taylor Black, a Microsoft AI executive who directs Catholic University of America's AI institute, said it will push people at the forefront of these tools to ask what it means to be human. Leo also used the document to issue the first papal apology for the Holy See's role in legitimizing slavery, by giving European sovereigns authority to enslave "infidels." What to watch now is how policymakers, and the Trump administration in particular, respond to a pope calling for the regulation they have been dismantling.