The Strait of Hormuz has exposed an uncomfortable truth about modern warfare: a military can destroy much of an adversary’s force, achieve overwhelming air superiority and strike thousands of targets, yet still fail to control the geography that gives the adversary strategic leverage.
Since the war with Iran began in late February, the United States and Israel have applied extraordinary military power. According to US government figures, Operation Epic Fury involved more than 10,000 sorties and strikes on more than 13,000 targets, including command-and-control nodes, air defenses, missile forces, drone infrastructure and naval assets. Washington later claimed that roughly 150 Iranian vessels had been destroyed. Even allowing for uncertainty in those assessments, the scale of the kinetic achievement is clear.
Yet by mid-August, months after the opening strikes, one of the war’s most important strategic questions remained unresolved: who actually controls the Strait of Hormuz? Last Friday, only two commercial vessels were observed transiting the strait during the day, compared with more than 130 a day before the war. Commercial vessels were still being attacked, and the United States remained compelled to maintain substantial naval and air power in the area. Airpower did not fail; it largely did what it was designed to do. The problem was that what remained after the strikes was a different kind of military problem.
Hormuz connects the Persian Gulf to the Gulf of Oman and the Indian Ocean. At its narrowest point, it is roughly 21 miles, or 34 kilometers, wide, but the effective maneuver space for large commercial vessels is much smaller: two shipping lanes approximately two miles wide. Before the war, roughly 20 million barrels of oil passed through the strait every day, alongside a major share of global LNG trade. Yet its military importance comes from its geometry. Iran controls the northern coastline, while Bandar Abbas, Qeshm, Larak, Hormuz Island and the Tunbs form a network of observation points, launch sites, concealed bases and staging areas overlooking a compressed maritime environment. For Tehran, Hormuz is not merely a waterway. It is a geographic weapon system.
Iran spent decades building a force designed to exploit that geography. It never tried to match the United States carrier for carrier or destroyer for destroyer. Instead, it built a system for sea denial. Sea control means being able to use an area while preventing an adversary from using it; sea denial requires much less. It is enough to make an opponent’s use dangerous, expensive and unpredictable. Iran does not need to sink an American aircraft carrier to succeed. It does not even need to physically close Hormuz. It only needs to make tanker owners, insurers and captains ask whether the risk is worth it.
Its force structure follows from that logic: naval mines, fast attack craft, mobile coastal anti-ship missiles, one-way attack drones, surveillance networks, small submarines and unmanned systems operating above, on and beneath the surface. None is decisive alone. The difficulty emerges when all of them exist simultaneously, dispersed across coastline, islands, tunnels and mobile positions. The adversary does not need to be everywhere; it needs to make you believe it could be anywhere.
A single mine not found can affect commercial traffic more than dozens already destroyed. One mobile launcher whose position is unknown can turn large areas of coastline into potential firing zones. This is asymmetric warfare at its most effective: turning inexpensive, distributed and difficult-to-find systems into an enduring problem for a technologically superior force.
Here lies the central paradox of the campaign. The United States and Israel proved they could destroy ships, missile batteries, radars, command centers and production sites at scale. But Hormuz exposed the boundary between destroying military capability and controlling operational space. After the initial campaign, the United States still had to launch a separate effort to restore freedom of navigation. Mine-clearance operations began in April, and in early May Project Freedom was launched to maintain safe passage for commercial shipping. More than 15,000 US service members were reportedly involved, supported by destroyers, helicopters and more than 100 manned and unmanned aircraft providing round-the-clock coverage. Meanwhile, more than 1,500 commercial vessels were waiting inside the Gulf, unable or unwilling to leave.
After one of the region’s most intensive air campaigns, more than 15,000 personnel and hundreds of platforms were still needed to secure passage through a relatively narrow body of water. Not because the United States lacked power, but because the mission had changed. Destruction is episodic. Control is persistent. A strike happens at a moment in time; control must exist every moment.
It is tempting to conclude that the West simply underestimated Iran. That is only partly true. The United States knew about these threats; they had appeared for years in intelligence assessments, war games, and CENTCOM planning. The deeper problem is conceptual: Western force design remains far better optimized for finding and destroying targets than for persistently understanding and controlling complex terrain. Modern militaries excel at the kill chain – detect, classify, decide, strike. Hormuz requires something broader: sense, understand, predict, persist, act, verify and repeat, thousands of times, every hour, for weeks or months.
Iran’s objective was never necessarily to survive as a conventional navy. It was to survive as a network. Networks regenerate, relocate and hide. A launcher on an island can be destroyed, but another can appear tomorrow. A shipping lane can be cleared of mines, but new mines can be laid at night. The question therefore eventually becomes not only how to destroy another launcher, but who controls the terrain that allows it to return. The United States has understandably avoided seizing strategic islands or parts of the Iranian coast because such operations carry major risks and escalation. But the cost of occupying terrain does not make the military problem created by that terrain disappear.
Geography still matters.
The answer is not a return to mass armies. Hormuz points toward a different force architecture. In environments that demand continuous presence across dangerous and contested spaces, humans should no longer perform every surveillance, patrol, detection and response mission. The persistent layer of the battlefield should increasingly be held by distributed networks of unmanned systems in the air, on the surface, underwater and on land.
But adding unmanned platforms is not enough. Hundreds of drones streaming more video into a command center merely move the bottleneck from the battlefield to the operator. The real challenge is to build systems capable of understanding the terrain they observe: connecting sensors, identifying abnormal behavior, detecting change, distinguishing civilian activity from emerging threats, reacquiring mobile launchers and operating when communications are degraded or denied.
This is the difference between automation, autonomy and cognition. In such an architecture, unmanned systems should increasingly own persistence – the continuous, dangerous and exhausting layer of military operations. Humans should be preserved for missions where human presence remains indispensable: strategic decisions, judgment under ambiguity, interaction with populations, seizure of critical infrastructure and, when necessary, physical control of terrain. Machines should own the persistence; humans should own the consequence.
Hormuz is an almost perfect laboratory for this model. A cognitive unmanned network could persistently monitor islands, coastlines, shipping lanes and the subsurface domain, detect and isolate threats as they emerge, and create the operational envelope within which manned forces enter only when a mission requires physical seizure or control of a critical point. Large-scale autonomy and manned ground forces are not alternatives. They are complementary: the first creates persistence; the second creates physical control.
The lesson of Hormuz is not that airpower failed. It is that airpower can succeed spectacularly and the strategic problem can still survive. Iran demonstrated how a conventionally weaker state can combine geography, inexpensive systems, and persistent uncertainty to generate global leverage. Future campaigns will therefore require militaries to move beyond a force architecture optimized mainly to destroy faster and toward one capable of understanding and holding continuously.
A military can destroy a navy, strike thousands of targets, and achieve overwhelming superiority in the air. But after the smoke clears, the most important strategic question may still be remarkably simple: who holds the ground? Hormuz is not an exception to the wars of the future. It may be their preview.
Ryan Gity authored Israel’s National Artificial Intelligence Strategy and served as a member of the team appointed by the Prime Minister to formulate Israel’s new defense force-building concept and future IDF budget framework. He is the CEO of G2, a deep-tech AI company focused on cognitive systems and physical AI.