THE STRAIT OF HORMUZ is twenty-one miles wide at its narrowest point, fringed by Iranian missile batteries, and — as of mid-March 2026 — functionally closed. Since the IRGC declared the waterway shut on March 2, tanker traffic has collapsed to near zero, more than 150 vessels sit at anchor in the open Gulf, and Brent crude has surged past $100 a barrel from a pre-war price of roughly $65. Some 20% of the world's daily oil supply is bottled up behind what analysts at the Center for Naval Analyses are calling an Iranian "kill box." The disruption, according to the International Energy Agency, is the largest in the history of the global oil market.
What makes this crisis especially galling is that the U.S. Navy spent more than four years preparing for something quite like it — and appears to have left the preparation on the shelf. In September 2021, Naval Forces Central Command stood up Task Force 59, a fledgling unit based in Bahrain with an explicit mandate to integrate unmanned aerial, surface, and subsurface systems into 5th Fleet operations across precisely this battlespace. TF59 reached full operational capability by January 2023. It tested some fifteen types of drones, accumulated over 55,000 operating hours, and participated in more than thirty multilateral exercises and six operational deployments around the Arabian Peninsula. Its crowning achievement: a complex manned-unmanned teaming operation conducted in the Strait of Hormuz itself. By December 2025 — just weeks before war broke out — TF59 had overseen the Navy's first-ever launch of a one-way attack drone, the LUCAS system, from a surface combatant in the Arabian Gulf. A dedicated drone strike squadron, Task Force Scorpion Strike, was deployed to the Middle East that same month.
The startups were there, too. Saildrone, a San Francisco-based company that builds wind-and-solar-powered unmanned surface vessels, has been embedded with TF59 since its inception. Its 23-foot Explorer drone — essentially a tall-winged ocean sailboat bristling with cameras, radar, and AIS transceivers — can stay at sea for up to twelve months without refueling, beaming surveillance data back to shore via satellite. One guided-missile destroyer costs roughly the same as 2,000 Saildrones; the economics of persistent maritime awareness, on paper, are overwhelming. Iranian forces twice attempted to capture Saildrone vessels in the Persian Gulf — a backhanded compliment to their operational value. The company also developed GPS-denial countermeasures after hostile actors tried to jam and spoof its vessels in the Red Sea. Beneath the surface, Palmer Luckey's Anduril Industries has been building what may be the most consequential unmanned naval platform of the decade: the Ghost Shark, an extra-large autonomous underwater vehicle developed with Australia's Defence Science and Technology Group. The Royal Australian Navy took delivery of its first production unit in January 2026 under a A$1.7 billion contract; days ago, on March 13, the Defense Innovation Unit and U.S. Navy selected Anduril's Dive-XL — Ghost Shark's commercial cousin — for the Combat Autonomous Maritime Platform project, a programme aimed at fielding long-range robotic submarines capable of mine detection, undersea surveillance, and operating in GPS-denied environments. Anduril's autonomous undersea vehicles have already logged more than 42,000 kilometers and 6,700 hours of mission time. The platforms exist. The production lines exist. The operational need is staring at the Navy from the front page.
Unmanned and unready
Yet when Iran's blockade materialized, the conversation in Washington reverted immediately to destroyers, carrier strike groups, and the operational mathematics of the 1987 Tanker War — not mesh-networked Saildrones and AI-enabled surveillance swarms. Senator Chris Murphy's post-briefing assessment was blunt: on the Strait of Hormuz, the administration "had NO PLAN." CNN reported that Pentagon and National Security Council officials significantly underestimated Iran's willingness to close the waterway, having assumed the economic self-harm would deter Tehran. Military officials privately told energy industry representatives that they could not spare Navy vessels for escort duty because those ships were already committed to offensive operations elsewhere in Operation Epic Fury.
The irony is thick. TF59 was built to be cheap, fast, and expendable — a "startup mentality," as its second commander, Captain Colin Corridan, put it — explicitly because the Persian Gulf's three chokepoints demanded persistent surveillance that billion-dollar destroyers could not sustainably provide. The unit's Saildrone Explorers and Devil Ray T-38 unmanned surface vessels could create mesh communication networks at sea, detect anomalous behavior through AI, and push maritime domain awareness into waters where manned patrols were too expensive or too risky to maintain. Bryan Clark of the Hudson Institute noted that TF59's greatest success was proving that commercial, off-the-shelf unmanned vessels could operate usefully in one of the world's harshest maritime environments. The task force even sent the first unmanned surface vessel through the Strait of Hormuz in April 2023, escorted by Coast Guard cutters, under the watchful eyes of Iranian observers.
Still, a demonstrable proof of concept is not the same as an operational capability at scale. Defence analysts had flagged the gap for years. Clark himself warned that TF59's contractor-owned model was ill-suited to military missions and that the Navy would need to take ownership of the platforms to employ them in combat. The bifurcation he hoped for — ISR missions shifting to regional partners while the Navy transformed drones for warfighting — appears never to have happened at the necessary pace. The result is a fleet that can launch a LUCAS drone from an LCS flight deck in a press-friendly demonstration but cannot, when the moment arrives, flood the Strait with low-cost unmanned scouts and strike assets to clear mines, map threats, and escort commercial shipping without risking billion-dollar warships and their crews.
The scale problem is damning. Analysts at Lloyd's List Intelligence estimate that a basic naval escort operation would require eight to ten destroyers to protect convoys of five to ten commercial vessels per transit — a commitment that could restore perhaps 10% of pre-war traffic. The Navy has roughly fifty combat-ready destroyers globally, dispersed across every theater. Jeff Currie, chief strategy officer at Carlyle Energy Pathways, told The Economist that the cost of a single manned escort would exceed the value of the cargo it was protecting. Meanwhile, Iran's asymmetric toolkit — drones, fast-attack boats, explosive USVs, and a mine stockpile estimated at 2,000 to 6,000 — is purpose-built for the narrow, shallow geography that TF59 was purpose-built to contest.
The Pentagon's predicament exposes a persistent dysfunction in American defence innovation: the valley of death between prototype and program of record, between a scrappy Bahrain-based task force and a fleet architecture that still counts hulls in the water. Four years of experimentation produced impressive operating hours and splashy exercises, but not the scaled, government-owned, combat-ready unmanned fleet that the moment demanded. The question now — with Brent hovering above $100, over 300 ships stranded in the Gulf, and allied nations declining to contribute warships to a war they opposed — is whether the Navy will finally treat unmanned systems as a mainline capability rather than an innovation side project. The robots were ready. The institution, it seems, was not. ■
For more, join 75,000 subscribers getting tech's favorite brief here
