The Illusion of Control: Where is the Human in Algorithmic Warfare?

The 21st-century battlefield is transforming at a dizzying pace, driven by the unstoppable march of artificial intelligence. What was once the stuff of science fiction is now a stark reality dominating headlines and crisis rooms. The recent legal dispute between Anthropic and the Pentagon, along with AI's increasingly prominent role in current conflicts like that in Iran, underscores an undeniable truth: AI is no longer merely an analytical tool. It has become an active player, generating real-time targets, coordinating missile interceptions, and guiding lethal swarms of autonomous drones.

Amidst this revolution, public and strategic conversation has focused on the need to keep “humans in the loop” (human in the loop). Pentagon guidelines, for example, posit that human oversight offers accountability, context, and nuance, while mitigating the risk of cyberattacks. However, this premise, as reassuring as it may seem, is a dangerous distraction. The imminent threat is not that machines will act without human supervision; the real crisis is that human supervisors have a limited, if not zero, understanding of what the machine is actually doing. The idea of a “human in the loop” in AI-driven warfare is, in essence, an illusion.

The Silent Evolution of AI in Armed Conflict

For decades, AI in the military sphere was primarily limited to data processing and intelligence. It analyzed vast amounts of information to identify patterns, predict enemy movements, or improve logistics. It was a support tool, an extension of human cognitive capacity. However, this phase has become obsolete. Modern AI has transcended its auxiliary role to become a direct participant in lethal decision-making and the execution of actions on the battlefield.

• Real-time target generation: AI systems are now capable of processing data from multiple sensors (satellites, drones, ground intelligence) and, with speed and precision unattainable by humans, identifying and prioritizing targets. They not only suggest but can designate and present attack options with algorithmic efficiency. This drastically reduces the time between detection and decision, but also compresses the space for human deliberation.
• Control and coordination of missile interceptions: In air defense scenarios, where every millisecond counts, AI is taking the reins. It can detect threats, calculate trajectories, determine the best response, and coordinate the launch of interceptors with perfect synchronization, surpassing any human reaction capability. The complexity and speed of these systems make human intervention almost symbolic.
• Guidance of autonomous drone swarms: Drone swarms represent a new frontier in warfare. Operating in a coordinated manner, these systems can saturate enemy defenses, conduct reconnaissance, targeted attacks, or even suppression missions. AI is the brain that orchestrates these swarms, adapting to changing battlefield conditions and making tactical decisions without constant human micromanagement. A human might give the order to deploy the swarm, but the execution and battlefield decisions are purely algorithmic.

The Mirage of Human Oversight

Given this reality, the notion of “humans in the loop” becomes increasingly difficult to sustain. It is not a matter of ill will or lack of ethics on the part of developers or the military, but an inherent consequence of the nature of advanced AI and the dynamics of modern warfare.

AI systems, especially those based on deep neural networks and machine learning, are notoriously opaque. They are known as “black boxes” because, although they produce impressive results, the internal process by which they arrive at those conclusions is extraordinarily complex and often inscrutable even to their creators. How can a human supervise, let alone be held accountable for, a decision made by an entity whose underlying reasoning is inaccessible?

Speed is another critical factor. Modern warfare is fought on a time scale that exceeds human processing and reaction capabilities. When an AI generates real-time targets or coordinates missile defense in fractions of a second, human intervention is not only slow but can be counterproductive. A human operator attempting to understand the context, verify information, and make a decision within milliseconds faces an impossible task. In practice, the “human in the loop” becomes a “human out of the loop,” or, at best, a “human in the approval loop,” where time pressure forces passive acceptance of AI recommendations.

Furthermore, cognitive fatigue and information overload are serious problems. Human operators are already overwhelmed by the amount of data they must process in a combat environment. Adding the task of monitoring and understanding the decisions of complex AI systems only exacerbates this burden, leading to errors, poor oversight, or an excessive and uncritical reliance on machine decisions. The “human in the loop” might be physically present, but their ability to exercise meaningful oversight would be severely compromised.

Beyond “Intervention”: The True Threat

The true danger does not lie in machines acting without human supervision; it is that human supervisors have no idea what the machine is doing, how it arrived at its conclusions, or what the unforeseen consequences of its actions might be. This lack of knowledge creates a false sense of security and accountability. Current guidelines, though well-intentioned, seem to be based on an outdated paradigm of human-machine interaction, where AI is a transparent and controllable assistant.

When an AI system fails or makes a mistake, the opacity of its operation makes it almost impossible to identify the root cause, learn from it, or assign responsibility. Who is to blame when an algorithm decides on a wrong target or a disproportionate action? The programmer? The operator who approved the decision without understanding it? The machine itself? This ethical and legal ambiguity is a ticking time bomb that threatens to undermine the principles of just war and accountability.

Ethical and Geopolitical Implications

The illusion of the “human in the loop” has profound ethical and geopolitical implications. If humans cannot understand AI's decisions, accountability dilutes to the point of disappearing. This opens the door to a dehumanization of warfare, where life-or-death decisions are made by algorithms, without the capacity for empathy, moral judgment, or contextual understanding that only a human can provide.

Furthermore, the AI arms race is accelerating, and nations that prioritize algorithmic speed and efficiency over meaningful human understanding and oversight could gain a short-term tactical advantage. However, this could lead to uncontrollable escalation, where conflicts unfold at algorithmic speeds, leaving little to no room for diplomacy or de-escalation. The unpredictability of AI systems could generate catastrophic conflict scenarios, where the actions of a machine trigger chain reactions that escape human control.

Conclusion: Awakening from the Illusion

The debate about “humans in the loop” is a comfortable distraction that prevents us from confronting the true and pressing question: how can we ensure that humans maintain meaningful control and deep understanding over the AI systems that are redefining warfare? The solution is not simply to demand the presence of a human; it is to develop AI systems that are more transparent, explainable, and auditable, and to establish robust legal and ethical frameworks that address the opacity and speed of algorithmic warfare.

It is imperative that the international community, governments, and AI technology developers set aside the illusion of the “human in the loop” and begin a more honest and urgent conversation about how to govern and understand these powerful systems. Only then can we aspire to a future where AI serves humanity without undermining the fundamental principles of responsibility, ethics, and meaningful control over the destiny of war. Algorithmic warfare is already here, and it's time for our understanding and policies to catch up with its relentless advance.