Pope Leo XIV stood before students and academics at Rome’s La Sapienza University on May 14, 2026, and said something that no previous pope had said from that stage. He said that AI warfare was pulling humanity toward annihilation. The speech was the first papal visit to the university since 2008, when Pope Benedict XVI cancelled a planned appearance amid faculty protests. Leo was welcomed warmly. What he said, however, was not comfortable.
“What is happening in Ukraine, in Gaza and the Palestinian territories, in Lebanon, and in Iran illustrates the inhuman evolution of the relationship between war and new technologies in a spiral of annihilation,” the American-born pope told his audience. He condemned the surge in global military spending as draining investment from education and healthcare while “enriching elites who care nothing for the common good.” He called for immediate and meaningful oversight of how AI is developed and deployed in military contexts, warning that the technology must not be allowed to “absolve humans of responsibility for their choices.”
The speech was remarkable for its specificity and for its timing. It arrived at a moment when the data on global military spending, AI weaponization, and the pace of autonomous weapons development all point in the same direction Leo described.
The Numbers Behind the Warning
Pope Leo’s condemnation of rising military spending was not a rhetorical abstraction. The Stockholm International Peace Research Institute released its annual military expenditure report on April 27, 2026, and the figures it contains are striking by any measure.
Global military spending reached a record $2.887 trillion in 2025, the eleventh consecutive year of growth. Spending has increased 41% over the past decade. Europe recorded a 14% surge to $864 billion, the highest level SIPRI has ever recorded for the continent and the fastest annual increase for NATO’s European members since 1953. Germany crossed the 2% of GDP threshold for the first time since 1990, with spending rising 24% to $114 billion, and has since pledged to reach 3.5% by 2029. Spain’s military budget leaped 50%. Poland devoted 4.5% of its GDP to defense, the highest share among all NATO members.
The global military burden now stands at 2.5% of world GDP, its highest level since 2009. In absolute terms, the world spent $352 per person on military activity in 2025. That is more than the annual income of hundreds of millions of people in the developing world. The full SIPRI Trends in World Military Expenditure report for 2025 provides country-level breakdowns and the methodology behind these figures.
Leo noted that this spending surge is occurring at the expense of education and healthcare, a claim that is empirically supported. The countries increasing their defense budgets most sharply are, in many cases, simultaneously restructuring public spending to accommodate those increases. The trade-offs are real and measurable.
What AI Warfare Actually Looks Like in 2026
When Pope Leo warned about AI in military contexts, he was not speaking about a hypothetical future. He was speaking about a present that is already operating across multiple active conflict zones.
Autonomous drone systems are now used in Ukraine, Gaza, Lebanon, and the broader Middle East conflict that has drawn in Iran. These systems can identify, track, and in some configurations engage targets with minimal or no real-time human oversight. The speed at which they operate frequently exceeds the decision-making capacity of human commanders in the loop, which means that meaningful human control over individual lethal decisions is becoming a practical fiction rather than a regulatory assurance.
In Ukraine, both sides have deployed AI-enabled drone swarms capable of coordinating attack patterns across multiple targets simultaneously. In Gaza, AI-assisted targeting systems have been used to accelerate strike decisions in a conflict where the density of the civilian population raises profound questions about accountability and discrimination, the two fundamental requirements of lawful targeting under international humanitarian law. The systems do not make those ethical distinctions automatically. They make them according to parameters set by humans, but at speeds and scales that make human review of individual decisions operationally impossible.
This is precisely what Leo meant when he said AI must not be allowed to “absolve humans of responsibility for their choices.” The architecture of autonomous weapons systems can create accountability gaps — situations in which a lethal outcome occurs but no individual human made a specific decision to cause it. The legal and moral frameworks that govern armed conflict were built around the assumption of human agency at the moment of lethal decision. AI warfare challenges that assumption at a structural level, not just at the margin. The emergence of autonomous weapons systems represents the same kind of competitive pressure dynamic that has historically driven arms races into territory where AI-enabled threats reshape the entire framework within which security and governance operate.
The Encyclical That Has Not Yet Arrived
Pope Leo identified AI as one of the most critical matters facing humanity immediately upon his election. He has said it publicly and repeatedly. The speech at La Sapienza was not an improvised comment but a carefully developed position that he is expected to articulate more fully in his first encyclical, due to be released in the coming weeks.
An encyclical is the most authoritative form of teaching document a pope can issue. It carries weight not just for the 1.4 billion Catholics who recognize papal authority but as a formal statement of moral philosophy that enters public discourse at a level that few other institutional voices can match. When John Paul II’s 1995 encyclical Evangelium Vitae addressed the sanctity of human life, it shaped policy debates in multiple countries and influenced legislation in ways that persisted for decades. Leo’s encyclical on AI is expected to do the same for a question that is currently being addressed almost entirely by technologists, military strategists, and regulators, with minimal input from the moral traditions that have historically set the outer limits of permissible conduct in warfare.
The specific call in the La Sapienza speech was for “better monitoring of how AI was being developed and used in military and civilian contexts.” That is a demand for governance, not a demand to stop AI development, but a demand to ensure that humans remain accountable for what it does. It is a position that is formally aligned with proposals being advanced by some governments and international bodies in ongoing negotiations over autonomous weapons standards, and that is being actively resisted by others, including the United States, Russia, and China, who have declined to support binding international restrictions on autonomous weapons development.
The full text of Pope Leo’s speech and its reception at La Sapienza University is covered in detail through the PBS NewsHour report, which includes the context of the papal visit and the student responses to the speech.
Why the Vatican’s Voice Matters Here
It is reasonable to ask what a pope’s speech at a Roman university has to do with the trajectory of AI weapons development. The answer is more substantive than it might initially appear.
Institutional religion remains one of the few forces in public life with genuine cross-cultural moral authority. The Catholic Church operates in virtually every country on earth. Its statements on questions of ethics and governance are reported, debated, and responded to by political leaders in ways that the statements of most other moral institutions are not. When Pope Francis issued Laudato Si in 2015, his encyclical on environmental ethics, it was credited by diplomats and policy analysts with contributing to the political atmosphere that produced the Paris Agreement later that year.
Leo is operating in a similar mode, but on a question that is moving faster and has fewer existing international frameworks to build on. The regulation of autonomous weapons systems is currently being discussed in UN forums under the Convention on Certain Conventional Weapons, a process that has been ongoing since 2014 without producing binding commitments. The major military powers have shown little appetite for restrictions that would constrain their own development programs. Into that governance vacuum, Leo’s speech injects a moral authority that is difficult for political leaders to simply ignore, particularly in countries with large Catholic populations, including much of Europe, Latin America, and parts of Africa and Asia.
The speech also landed in a week of extraordinary geopolitical turbulence. The US-Iran conflict has expanded the geography of active hostilities in the Middle East. Ukraine is in its fourth year of a war that has become a laboratory for AI-enabled weapons systems on both sides. The competition for strategic position in contested spaces from the Arctic to orbital infrastructure is increasingly shaped by autonomous and AI-directed systems that operate with minimal human oversight. The race to dominate strategic orbital infrastructure is itself a dimension of the AI arms race Leo described, one in which the speed and autonomy of space-based systems create command and control challenges that existing governance frameworks were not designed to address.
The Part of the Speech Nobody Quoted
The headlines from Leo’s La Sapienza speech focused, understandably, on the “spiral of annihilation” phrase. But the section of the speech that may prove more durable was quieter and less quotable.
Speaking directly to the students, Leo said: “We are a desire, not an algorithm.” He urged them not to give in to resignation but to “transform restlessness into prophecy.” He told them that education and research must move in a direction that values life rather than the accumulation of firepower or profit. And he told them that the crisis of the world they are inheriting is not primarily a technological crisis but a spiritual and moral one — a crisis of what human beings decide to value and what they decide to build.
That distinction matters because the most common frame for discussing AI in warfare is technical and legal: what systems are possible, what systems are permitted, and what verification mechanisms could be implemented. Leo’s frame is different. He is asking what kind of civilization chooses to invest $2.9 trillion per year in military activity while telling teachers and doctors there is no money available for the things they need. He is asking what it means for human responsibility and human dignity when machines are assigned the task of making decisions that end lives.
Those questions do not have technical answers. They are the kind of questions that have historically been the province of philosophy, theology, and ethics disciplines that have been largely absent from the most consequential conversations about AI development happening in laboratories, boardrooms, and government ministries. The geopolitical competition reshaping access to strategic resources reflects exactly the kind of logic Leo was challenging: the assumption that national advantage is the primary frame within which consequential decisions about technology and security should be made.
Looking Ahead
Pope Leo’s encyclical on AI, when it arrives, will be one of the most significant institutional statements on artificial intelligence issued by any organization in 2026. It will be scrutinized by governments, technology companies, military establishments, and civil society organizations across the world. It will not change policy by itself. Encyclicals never do.
What they do is alter the moral vocabulary available to people arguing for or against particular positions. They create legitimacy for concerns that were previously dismissed as sentimental or impractical. They make it harder for decision-makers to act as though the ethical dimensions of a question do not exist, because a recognized moral authority has placed them explicitly on the table.
The questions Leo raised at La Sapienza about accountability in autonomous weapons, about the trade-offs between military spending and human welfare, about whether AI should be allowed to make decisions that end human lives — are not going away. They are going to become more pressing as the systems become more capable, more widely deployed, and more deeply integrated into the infrastructure of conflict that is currently shaping outcomes in Ukraine, Gaza, and the broader Middle East.
“Let us not call defense a rearmament that increases tensions and insecurity,” Leo said, standing in a lecture hall founded by a pope in the fourteenth century, speaking to students from Gaza who had arrived in Rome two days earlier. “It drains investments in education and health, contradicts trust in diplomacy, and enriches elites who care nothing for the common good.”
The data support the analysis. The America Magazine coverage of the La Sapienza speech provides additional context on how Leo’s position on AI warfare fits within the broader trajectory of his papacy and his forthcoming encyclical. The question of what to do about it remains, as it usually does, a human one.
