M A Hossain,
There is a moment in the history of every transformative weapon when it stops being a tool and starts being a doctrine. The machine gun remade European civilization, burying a generation in Flanders mud. The atomic bomb ended one war and restructured global politics for generations — producing a permanent architecture of mutual dread we still inhabit. Each time, the technology moved first. Accountability scrambled after.
We are at that inflection point again. Artificial intelligence(AI) is not approaching the battlefield. It is already there — identifying targets, compressing decision cycles, reshaping the fundamental grammar of modern warfare. The question is no longer whether AI transforms how wars are fought. It does. The urgent, uncomfortable question is whether democratic societies ever actually 'chose' this — or simply woke up inside it.
Project Maven, the Pentagon's flagship AI targeting program, began modestly enough. Analysts drowning in drone footage needed relief. Let machine vision sort the imagery, the argument went, and free human judgment for what matters. Track motorcycles. Flag patterns. Reasonable.
But numbers have a way of growing. What processed a hundred targets daily without AI processed a thousand with it. Add large language models, officials suggested, and five thousand targets a day becomes plausible. Five thousand. The analyst once strained over fifty images is now nominally "in the loop" on decisions generated at the industrial scale. At that velocity, "human oversight" begins to resemble a public relations formula more than a genuine safeguard.
Jim Mattis put it plainly: targetry is not a strategy. Speed and scale are not wisdom. The United States spent punishing years in Iraq and Afghanistan learning exactly that — that the ability to strike something tells you nothing about whether striking it advances any coherent purpose. AI creates a powerful temptation to forget that lesson entirely.
The Iran operations illustrated this precisely. Central Command confirmed AI had compressed decision cycles from days to seconds. Impressive, technically. Yet the regime stands. No strategic realignment followed. Meanwhile, credible reports of civilian casualties — including a struck school — surfaced and then dissolved into procedural conversations about data labeling. The dead became a debugging problem. That is a profound moral failure dressed in technical language.
Ukraine offers a more instructive alternative. The Delta battlefield awareness platform achieved near-total situational awareness without surrendering targeting authority to autonomous systems. Human operators remained genuine decision-makers — not rubber stamps on machine-generated kill lists. Imperfect, certainly. But proof that the choice is not binary between algorithmic warfare and battlefield blindness.
The deeper crisis, however, is political. This transformation of American warfare has unfolded almost entirely outside democratic deliberation. No legislature debated it seriously. No public consented to wars waged at machine speed. The phrase "appropriate levels of human judgment" — the Pentagon's current formulation — was composed by officials, not authorized by citizens.
Technologies that reshape war consistently outpace the institutions meant to govern them. AI is moving faster than any predecessor. Who draws the line? So far — nobody in particular, moving very fast, toward a destination nobody formally chose. That should disturb us. Profoundly.
M A Hossain is a senior journalist and international affairs analyst, based in Bangladesh. He can be reached at: writetomahossain@gmail.com
This article published at :
1. The Korea Times, S Korea : 03 June, 26
