Pope Leo XIV’s warning on AI warfare comes as autonomous weapons move from theory to reality
Stop Killer Robots UK welcomes Pope Leo XIV’s strong and timely intervention on artificial intelligence and warfare. At a moment when AI-enabled targeting systems are moving from experimental capabilities to operational reality, the Pope’s encyclical (a formal papal letter addressed to the Church and wider world on major moral and social issues), Magnifica Humanitas, provides a moral framework for policymakers seeking to establish rules around the future of warfare.
The debate over military artificial intelligence is often framed as a contest between innovation and regulation. Pope Leo XIV’s first encyclical, Magnifica Humanitas, challenges that framing. His Holiness asks a more fundamental question. How much human responsibility can be removed from warfare before the use of force itself becomes morally unacceptable?
The Pope’s intervention is notable not because it rejects technology. On the contrary, the encyclical explicitly acknowledges that technological innovation can serve humanity and contribute to the common good. Rather, it argues that technology must remain subordinate to human dignity, moral agency and political accountability. Those principles have particular relevance as governments struggle to agree international rules governing autonomous weapons systems.
At the centre of the encyclical is a clear warning: “there is no algorithm that can make war morally acceptable.” This is more than a theological observation. It is a direct challenge to a growing trend in military planning that seeks to increase the speed, scale and autonomy of battlefield decision-making.
From Gaza to Iran
The principle of meaningful human control has become the cornerstone of international discussions on autonomous weapons systems. The concept reflects a simple but fundamental proposition: humans, not machines, should remain responsible for decisions to use force.
Recent conflicts demonstrate why this matters.
In Gaza, AI-enabled targeting systems known as “Lavender” and “Gospel” brought public attention to the role of algorithms in military targeting. According to multiple investigations, Lavender was used to rapidly identify individuals suspected of links to Hamas or Palestinian Islamic Jihad, while Gospel was designed to generate recommendations for buildings and infrastructure that could be considered military objectives. Critics argued that the systems accelerated targeting processes to a degree that risked reducing meaningful human review, while Israeli authorities maintained that human operators remained responsible for all strike decisions.
The controversy marked a significant shift in public understanding of military AI. The debate moved beyond fully autonomous weapons and towards a broader question: can human control remain meaningful when algorithms generate targets faster than human operators can realistically assess them?
A similar trend has emerged in relation to Iran. Analysts have highlighted the growing role of AI-assisted intelligence analysis, target identification and mission planning in recent military operations involving Iran, the USA, and Israel. While many of these systems remain decision-support tools rather than fully autonomous weapons, they illustrate how AI is increasingly shaping the targeting cycle itself. The concern is not necessarily that machines are independently deciding whom to kill; it is that human decision-makers may become increasingly dependent upon algorithmic recommendations that are difficult to challenge or scrutinise.
The Pope’s warning speaks directly to this challenge. Throughout Magnifica Humanitas, he rejects the notion that technological capability can substitute for moral responsibility. This principle becomes increasingly important as military organisations seek to compress decision-making timelines and operate at machine speed.
Ukraine and the acceleration of military AI
The urgency of this issue has been demonstrated by the conflict in Ukraine. Since the Russian invasion the use of novel technologies has characterised this conflict. What started with the use of commercially available drones has evolved into a testing ground for increasingly sophisticated AI-enabled systems. This has included technologies that are capable of automatic targeting and engagement to counter jamming.
Both Ukraine and Russia are investing heavily in technologies designed to operate in environments where communications are disrupted by electronic warfare. AI-assisted drones are increasingly capable of navigating, identifying targets and completing missions with reduced reliance on direct human control. Recent reporting suggests that Ukraine is deploying AI-enabled drones capable of autonomously identifying and tracking targets after launch, while both sides are experimenting with systems that can continue operating when communications links are jammed.
The trend extends beyond aerial systems. Ukraine has dramatically expanded its use of unmanned ground vehicles for logistics, casualty evacuation and combat support. The result has been that military planners are increasingly discussing the integration of autonomous systems across the battlefield.
Supporters argue that such technologies can save soldiers’ lives, compensate for manpower shortages and improve military effectiveness. These are serious arguments that cannot be dismissed, particularly in the context of a war of national survival. Yet Ukraine also demonstrates why governance questions can no longer be treated as hypothetical.
As AI systems assume greater roles in target identification, navigation and engagement, the distinction between automated assistance and autonomous decision-making becomes increasingly blurred. The trajectory of technological development points towards greater autonomy, not less.
The challenge of meaningful human control
This is where Pope Leo’s intervention intersects with one of the most important debates in international security policy: meaningful human control.
For more than a decade, governments, civil society organisations and experts have argued that humans must retain meaningful control over decisions to use force. The principle seeks to ensure that moral and legal responsibility cannot be delegated to machines. Human beings, not algorithms, should remain accountable for decisions involving life and death.
The Pope’s concern reflects the same underlying principle. Throughout the encyclical, he rejects the idea that technical capability creates moral legitimacy. “To disarm,” he writes, “does not mean rejecting technology, but preventing it from dominating humanity.”
The challenge is that meaningful human control is becoming harder to define as military AI systems grow more sophisticated. Modern systems may not independently choose targets in the science-fiction sense often imagined in public debate. Instead, they can shape targeting decisions, prioritise threats, recommend engagements, filter information and operate in communications-denied environments. Human operators may remain formally “in the loop” while exercising increasingly limited practical control over outcomes.
The question facing policymakers is therefore not whether humans remain involved, but whether their involvement remains meaningful.
Beyond efficiency and military advantage
The significance of Magnifica Humanitas lies in its rejection of a purely technocratic approach to AI governance. The encyclical warns repeatedly against allowing efficiency, power or competitive advantage to become the primary criteria by which technologies are judged.
This matters because discussions of military AI are often dominated by strategic considerations. States fear that restricting autonomous weapons could leave them vulnerable to adversaries willing to deploy them. The result is a classic security dilemma in which every actor feels compelled to pursue greater autonomy because others are doing the same.
Ukraine illustrates this dynamic vividly. Faced with an existential threat, Ukrainian innovators have pushed the boundaries of military AI to maintain battlefield effectiveness against a larger adversary. Russia has responded with its own investments in increasingly sophisticated drone technologies. The logic of competition drives both sides towards greater autonomy.
Pope Leo’s argument is that technological competition cannot be allowed to determine humanity’s moral boundaries. The existence of a capability does not automatically justify its deployment.
An intervention at a pivot point in international negotiations
The encyclical arrives at a moment for international discussions on autonomous weapons. Diplomatic efforts within the framework of the Convention on Certain Conventional Weapons have struggled to produce binding rules, despite broad agreement that some form of human responsibility must be preserved. Meanwhile, advances in AI-enabled military systems continue to outpace the speed of diplomacy and regulations.
Pope Leo’s intervention argument is not merely that AI poses risks. It is that human dignity requires limits on the delegation of lethal decision-making.
The lesson emerging from Ukraine is not that autonomy is inevitable. It is that the window for establishing meaningful safeguards may be closing. As AI systems become more capable and more deeply embedded within military operations, preserving meaningful human control becomes increasingly difficult.