The War Runs on Chokepoints and Guesswork

Six months into the US war with Iran, the central fact is less a battle plan than a plumbing problem. Tehran’s effective closure of the Strait of Hormuz has made the movement of oil—not merely the destruction of military targets—the organising principle of the conflict. Brent, which averaged roughly $69 a barrel last year, was around $100 at publication and had briefly hit $126 in late April. That is painful, but not the apocalypse once forecast. The distinction matters: a crisis that has not yet wrecked the global economy can still be steadily wrecking the political and strategic assumptions underneath it. The Associated Press account is here.

China has supplied an awkward part of the answer. Its strategic crude reserve was estimated by the US Energy Information Administration at about 1.4 billion barrels at the end of last year. After the bombardment began and Hormuz became effectively closed, Beijing drew down stocks and sharply reduced imports; US data put its second-quarter crude imports at 8.1 million barrels a day, nearly four million barrels a day below the prior quarter. China’s electric-vehicle shift and other alternatives also reduced demand. In effect, Beijing’s insurance policy cushioned a war begun by Washington and Israel. Geopolitics retains its capacity for bleak farce: the administration confronting China may partly owe its less-disastrous petrol-price problem to Chinese contingency planning. The price and stockpile details are reported here.

That buffer is not a settlement. Bank of America analysts projected markedly higher prices if Hormuz traffic remains restricted, and still worse spikes if major energy infrastructure is damaged. Meanwhile, attacks by Iran-backed militias prompted Saudi Arabia to temporarily shut a vital pipeline to Red Sea ports, proving that rerouting around one chokepoint can simply introduce another with more missiles. A war advertised in military terms is being adjudicated by tank farms, pipelines and maritime insurers—the least cinematic cast in international affairs, and often the one with the final word. Those risks and scenarios are detailed by AP.

The most alarming escalation described this week was one that did not happen. According to a Gizmodo report summarising CNN’s reporting, US forces planned to board a China-flagged vessel in the Middle East after a chatbot used by a special-operations analyst allegedly identified its cargo as material potentially usable for Iran’s nuclear programme. The boarding was called off. Details remain scarce: the reporting does not identify the vessel, chatbot, location, or the human review that prevented the operation. But a near-miss does not become trivial because the brakes worked at the last junction. The reported incident and its limitations are here.

Military institutions have always made errors; the novelty is the sales pitch that turns a text-prediction system into a confidence machine. An AI output can arrive with the syntactic poise of analysis while concealing the more prosaic fact that it may have confused, fabricated, or over-inferred. In a crowded regional war, wrongly boarding a Chinese-flagged ship is not an ordinary intelligence failure. It is a possible collision between the Iran conflict and US-China relations, with sailors and diplomats left to sort out what the software’s prose had made urgent.

The same report points to a separate Bloomberg-reported allegation that analysts relied too heavily on Palantir’s Maven Smart System before a US missile strike on Shajarah Tayyebeh Elementary School on the war’s first day. The article says the strike killed more than 150 people, including 123 children, and that UN experts found reasonable grounds to conclude war crimes may have been committed; that is not a final judicial determination. The ethical issue is not that software alone bears blame—humans make targeting decisions—but that automation can distribute responsibility so efficiently that nobody seems to be holding it when the consequences arrive. Gizmodo’s account attributes the school-strike reporting and notes the uncertainty.

Saudi Arabia’s first air-raid alerts for Riyadh since the Houthis escalated attacks in July are a reminder that no sea lane stays merely commercial once a regional war gets its hands on it. Reuters and AFP journalists heard explosions in the capital; authorities later issued an all-clear, though the available report does not establish what caused the blasts. The alerts followed Houthi attacks on Saudi oil facilities that caused fires and temporary operational halts. The BBC reports the alerts and the unresolved details.

The strategic geometry is almost offensively neat. With Hormuz effectively closed, Saudi vessels have relied on Red Sea routes for exports to Asia and Europe. The Houthis have seized Mokha and Perim Island, positions the BBC describes as gateways to a key Saudi shipping route, while declaring a maritime embargo on Saudi Arabia. The Houthis say their blockade answers Saudi restrictions on Houthi-controlled ports and airports; Saudi Arabia and Yemen’s government accuse them of choking the Bab al-Mandab to exert economic pressure. Whatever the competing justifications, oil cargoes are now travelling through a conflict whose local roots predate this wider war by more than a decade. The relevant maritime and political claims are set out here.

That history should puncture any temptation to treat Yemen as a convenient side quest. Its civil war began in 2014 and escalated when a Saudi-led coalition intervened in 2015; the BBC reports more than 22 million people need some form of aid, citing the UN. Two sources told CBS that Crown Prince Mohammed bin Salman asked President Trump for direct US military action against the Houthis, and that Trump declined while offering intelligence and targeting support. If accurate, that is restraint of a particularly conditional sort: America may not add its own aircraft, but it remains available to improve the coordinates. The BBC includes the sourcing and caveats on that account.

Washington’s new Russia sanctions law demonstrates how quickly one war’s economic toolkit gets wheeled toward another. The legislation targets Russia’s energy and defence sectors, senior officials, the tanker shadow fleet, and entities supporting Moscow’s military; it also extends sanctions involving Iran. Crucially, it authorises tariffs of up to 100% on major buyers of Russian oil and gas, potentially including China and India. Authorisation is not implementation, and the supplied reporting does not say these tariffs have been imposed. Still, the threat places two of the world’s largest energy consumers inside the same coercive frame as a Middle Eastern oil crisis. AFP’s report on the signed law is here.

The timing is less than elegant. India said it would continue buying oil through diversified sourcing, while China said it reserved the right to protect its sovereignty, development interests and companies’ rights. Trump signed the measure a week before a planned Washington meeting with Xi Jinping. Meanwhile, China’s own stockpile drawdown has helped moderate the oil shock from the Iran war. Washington is therefore asking Beijing to be, depending on the hour, a strategic competitor, a prospective tariff target, a diplomatic interlocutor on Iran, and an involuntary stabiliser of the global energy market. Empires have managed stranger arrangements. They do not generally make them less combustible. The sanctions reporting records the prospective scope and responses.

The connecting lesson is not that economic pressure is useless. It is that it is now being deployed in a world where crude reserves, sanctioned tankers, militia-held islands, shipping chokepoints and questionable machine-generated intelligence all interact. A tariff can punish a buyer; it cannot reopen Hormuz. A naval patrol can protect a route; it cannot erase Yemen’s war. An AI tool can accelerate an assessment; it cannot inherit judgment. The conflict’s true escalation ladder is not made of rungs but supply chains, and every state seems determined to climb it while carrying gasoline. The oil-market context is here.