How Trump hacked the presidency

By Jason Willick. 

Like Anthropic's Mythos AI model, the president has a supercharged ability to find a system’s weaknesses.

May 31, 2026 

 Think of Donald Trump as the Mythos president. Anthropic’s Mythos artificial intelligence model created a media sensation for its purported ability to find and exploit vulnerabilities in thousands of computer systems. It didn’t create vulnerabilities — they were there all along and occasionally caused problems. But Mythos apparently has a special knack for smoking them out.

Trump has a similar ability for the presidency. He has cleverly found and exploited ambiguities in the office to circumvent traditional limits. His lawsuit against his own government, and the resulting “settlement” involving an extrajudicial payout fund for allies, is a prime example.

That lawsuit took advantage of the fact that the president is “the only person who alone composes a branch of government,” as the Supreme Court put it in 2020. There are more than 300 million American citizens who can sue the federal government to right a perceived wrong. The president is the only one among them who, if he does so, is actually suing himself.

Trump was wronged by the leak of his tax returns in 2019 and 2020. It wasn’t out of bounds for him to sue the IRS over the breach. But for Trump’s “public” self to settle a suit with his “private” self? That doesn’t compute.

The administration didn’t hesitate to exploit the vulnerability. The Justice Department announced its intent to “settle” Trump’s suit with the creation of a $1.776 billion “Anti-Weaponization Fund.” The idea is to pay people the administration thinks have been mistreated by the government, most likely the president’s political allies such as participants in the Jan. 6, 2021, Capitol riot.

It bears repeating that Trump is pressing on a vulnerability that already existed, not blowing up the system outright. As the University of Chicago law professor Aziz Huq writes, past administrations have used federal settlements to generate payouts “to politically preferred groups even when there was no real prospect of a legal challenge against the government prevailing in court.”

Huq cites as one example the Clinton administration’s monetary settlement with Japanese people from Latin America whom the United States interned during World War II, even though the law only provided for payments to Japanese who were U.S. citizens or residents. “There is, to be sure, no moral equivalence between the interned Japanese and the Jan. 6 rioters,” Huq writes.

That statement is so obviously true it’s jarring. But it’s essential to understanding Trump’s Mythos quality. He can look at a system coldly, without regard for moral judgments. If a vulnerability exists, then it can be exploited to benefit morally deserving and undeserving groups alike. A Democratic-appointed federal judge on Friday launched an inquiry into the settlement, but that seems to reflect a moral objection rather than a strictly legal one.

Another part of the settlement claims to shield Trump from certain future legal actions by the U.S. government, including IRS audits based on his already-filed tax returns. Scholars debate whether a president can pardon himself, but in some ways this deal is broader, because a pardon applies only to criminal charges. The agreement, signed by acting attorney general Todd Blanche, covers civil fines as well.

Once again, the president is exploiting a bug in the law that was not entirely unknown. When the Biden Justice Department was prosecuting Trump in 2024, it tried to persuade the Supreme Court that former presidents didn’t need broad immunity from prosecution. After all, a lawyer for the Biden administration told the justices, a president is already shielded from prosecution for doing things his attorney general tells him are legal.

The idea is that the attorney general represents the U.S. government, and the U.S. government can’t go back on its word to anyone — even the president. Now the U.S. government has formally agreed to limit future Trump investigations. But when the president is the government, the whole theory short-circuits, and Trump is happy to benefit from the ensuing confusion.

Another “hack” the administration may be perfecting is the use of indictments of foreign leaders to justify military interventions. The indictment, unsealed last week, of Cuban strongman Raúl Castro over the 1996 shoot-down of two planes wasn’t really a law enforcement action; it was a military threat. After all, the administration took care to describe its January military incursion into Venezuela to arrest dictator Nicolás Maduro and take greater control of the country’s politics as a “law enforcement” operation pursuant to a criminal indictment for drug and gun trafficking. Now it’s signaling Castro could be next.

The Constitution intended for Congress to play a key role in decisions to award government payouts and launch wars. That role has been eroding for some time, but Trump has been particularly cunning about finding vulnerabilities. Want a taxpayer fund for allies? Just sue the government in your private capacity and order your Justice Department to settle. Want to send U.S. forces to a foreign capital to kill its security forces and seize its ruler? Just get a grand jury in New York to rubber-stamp an indictment of the tyrant.

Mythos’s hyper-competence at finding software weaknesses has prompted a frenzy of bug-patching in the software that makes the world run. The executive branch needs some bug-patching, too.

By Jason Willick

Jason Willick is a Washington Post columnist focusing on law, politics and foreign policy.follow on X@jawillick

Harvard vs. Trump is the result of a massive error

The university argues its scientists are a public good. It needs politics to make that case.

The Washington Post. 

May 27, 2026.

6 min

By Harvey Mansfield

Harvey Mansfield is a retired professor of government at Harvard University and the author of “Where Harvard Went Wrong,” from which this op-ed, published originally in the Harvard Crimson, is adapted.

Lack of viewpoint diversity is the main field of the battle still raging between Harvard and the Trump administration, which in March sued the university for allegedly failing to protect Jewish students from discrimination and sought to halt its federal grant payments. Until recently, “diversity” referred to race and sex, but opponents of this sort of diversity argued that the term should include opinion as well. “Viewpoint diversity” is their term. Clearly diversity is the darling of the left, and viewpoint diversity the counter from the right.

Harvard’s one-sided fondness for the left, comprehensive and prolonged, provoked its clash with the Trump administration. It also revealed a deeper division between science and the humanities — quiet now, but with a Harvard history.

Viewpoint diversity means wanting more conservatives, not a further sprinkling of garish extremes. To lack a proper mix of left and right is not legally a crime — a reasonable point for Harvard — but it is a massive error, one that has forced Harvard into the courts. To depend on the courts to defend the university’s independence from federal oversight is still dependence, and it offers only tenuous relief from a Trumpist siege. One can see the risk in depending on courts within the very policy of diversity, when that word was suggested to universities in Regents of the University of California v. Bakke, the Supreme Court’s 1978 case on affirmative action in student admissions. Yet in 2023, the court reversed itself, stymieing the universities by declaring race-based affirmative action in admissions illegal and unconstitutional. At this point, “viewpoint diversity” redeems diversity only by calling for the inclusion of diversity’s opponents.

If Harvard wants to prevent further trouble with Republicans, it needs to change its attitude. As the Harvard administration has begun to see, it needs to drop its gratuitous partisan posture. There is much to gain and little to lose in welcoming conservatives to the university.

Who risks the most in Harvard’s battle? The scientists, who need the government’s money to carry on their work of usually expensive experimentation in laboratories. This opens up a second problem of viewpoint diversity within universities that is not so easy to fix: the divide between the scientists and the humanists.

When I arrived as a freshman at Harvard, I took a course, Natural Sciences 4, taught by the university’s then-president James B. Conant. He was a scientist who took a leading part in the direction of the Manhattan Project, which produced the atomic bomb. Absorbed in the question of whether this was a good thing for humanity, he inspired a new program at Harvard called “General Education,” to which this course belonged. To make education “general,” including both science and humanities, seemed to be the goal.

After this course, the differences between these fields became clearer to me. Science deals with numbers, the humanities with persons — the particular human beings from which science abstracts. If a course takes up proper nouns like Shakespeare and Goethe, it belongs to the humanities; if it concerns impersonal objects as common nouns, it is science.

Our names are essential to us. Every human being has a proper name, each different, to respect the human desire for self-importance. Yet a doctor using medical science does not need to know your name to treat your human body. Science is a nameless, collective enterprise of what “we know.” Humanities professors write their own books and make “contributions” to Shakespearean literature.

Science progresses by discarding old hypotheses and finding new ones. The humanities do not progress — who now equals Homer and Shakespeare? — but offer insights into permanent human truths of honor and beauty that the scientific method cannot discern or recognize.

Humanities pore over dusty books and archives. Science, with its “pioneering research,” can, by contrast, deliver manifest benefits, above all in modern medicine. But science or its technology also delivers risks to humanity from possible atomic warfare and climate change. Perhaps science has remedies for the dangers it brings, but perhaps not.

Moreover, science needs to address and convince the nonscientific public that its research is worth funding. This is difficult, because the exactness of modern science arises from its use of mathematics, which keeps it remote from the great majority of human beings who are not adept with numbers. The public has to be addressed with rhetoric — which is inexact and often promises too much. Experience shows that science itself is both hypothetical and open-minded, while the task of conveying science is often partisan and closed-minded.

Science dominates the university, but it cannot defend or explain itself without departing from scientific rigor. There is no scientific proof that science is good.

Yet turning from science to the humanities for assistance, one encounters postmodern arguments that flounder in helpless relativism. Far from giving reasons science might be good, the humanities fail to justify even themselves.

Why should Harvard be independent? Because it helps society, the university’s advocates say; it’s worth the money! But doesn’t the university need some standard from outside society to justify itself as independent? Something like Veritas (Harvard’s motto) that combines science and the humanities — a Harvard that looks for the wisdom that makes science valuable to human beings.

A political scientist myself, I like to think that this wisdom centers on politics. A wiser politics than devotion to a single party would have protected the scientists and corrected the humanists.

 

Otzio test for dialogue with speech marks

“Hi there Zack. You texted me?”

“Maya, hi. I did. You’ve been lying on me! You told me that Japanese curries are mild and inoffensive, when they’re not at all.”

“I wasn’t lying. Japanese curries ARE mild. You won’t find any milder in any other country, at least not in Asia. I can’t speak for Europe.”

“What about Coco Ichibanya?”

“Oh.”

“Exactly. You never told me about their spice level. ”

“Yes, you’re right! Completely slipped my mind. It goes from —”

“— from 1 to 20. We went there last weekend, to the franchise near our train station. I must have walked past it countless times and never once thought to go in.”

“I guess it’s got that family chain restaurant feel. I’m not a big fan of curry, but every now and then I tag along with my husband to Ichibanya. I never have anything too spicy — I suppose you’d say I don’t beat the stereotype! But he loves his curry.”

“What spice level does he order?”

“It’s usually in the high digits, never spicier than 10. And you? Where did you land?”

“I’m rather pedestrian, so I opted for 5. Joseph wanted to show off and try 15, but they wouldn’t let him!”

“What? Did they think he was a kid?”

“That wasn’t it. They said he had to have tried 10 first. They were willing to take his word for it, but he couldn’t lie; this was his first time at Coco Ichibanya. So he said, fine, I’ll try 10. But then they said he had to have tried 5 first!”

“No!”

“Yes! So in the end, we both had 5, which was kind of spicy but definitely left us wanting more. Next time, we qualify for 6 to 10.”

“I think I know where you’re going next weekend.”

Test for Otzio

The pageant has been a headlining event at Gathering of Nations, a massive and at times controversial event that bills itself as the largest powwow in North America, for more than four decades. Young women from across the U.S. and Canada competed for the prestigious title and iconic, intricately beaded crown.

Maya: Hi there Zack. You texted me?

Zack: I did. You’ve been lying on me! You told me that Japanese curries are mild and inoffensive, when they’re not at all.

Maya: I wasn’t lying. Japanese curries are mild. You won’t find any milder in any other country, at least not in Asia. I can’t speak for Europe.

Zack: What about Coco Ichibanya?

Maya: Oh.

Zack: Exactly. You never told me about their spice level.

Maya: Yes, you’re right! Completely slipped my mind. It goes from —

Zack: — from 1 to 20. We went there last weekend, to the franchise near our train station. I must have walked past it countless times and never once thought to go in.

Maya: I guess it’s got that family chain restaurant feel. I’m not a big fan of curry, but every now and then I tag along with my husband to Ichibanya. I never have anything too spicy — I suppose you’d say I don’t beat the stereotype! But he loves his curry.

Zack: What spice level does he order?

Maya: It’s usually in the high digits, never spicier than 10. And you? Where did you land?

Zack: I’m rather pedestrian, so I opted for 5. Joseph wanted to show off and try 15, but they wouldn’t let him!

Maya: What? Did they think he was a kid?

Zack That wasn’t it. They said he had to have tried 10 first. They were willing to take his word for it, but he couldn’t lie; this was his first time at Coco Ichibanya. So he said, fine, I’ll try 10. But then they said he had to have tried 5 first!

Maya: No!

Zack: Yes! So in the end, we both had 5, which was kind of spicy but definitely left us wanting more. Next time, we qualify for 6 to 10.

Maya I think I know where you’re going next weekend.

Oil markets are still in La La land

The Economist

Prices have risen sharply. Unfortunately, they still have further to go

Apr 30th 2026

|

6 min read

SoMEONE WAS sniffing the butane. Energy experts have long warned that the war in Iran was causing the biggest oil-supply shock in history. The closure of the Strait of Hormuz shut in 14m barrels a day of oil. To destroy that much demand, they said, the price of Brent crude should be more than double its pre-war level, at well over $150 a barrel. But oil traders were in a stupor. As recently as April 17th prices were below $90 a barrel. Over the past week, on talk of renewed fighting, they have been waking up. On April 30th prices spiked above $125.

Unfortunately, as bad as things are, the disconnect with reality endures. Not only may spot prices have further to climb, but the oil-futures market, in which speculators bet on where the oil price is going, says prices will fall every month for the rest of the year, ending 2026 at about $88. That implies most of this shock will soon be reversed. If so, traders must believe three things are true: that America and Iran will soon strike a peace deal; that their agreement will reopen Hormuz; and that, soon after the strait is clear, petrol and jet fuel will once again be plentiful. All those are in doubt.

One thing everybody should be able to agree on is that for the strait to stay shut would be a disaster. At the start of the war, lots of oil was in stocks or tankers at sea. But the ships that passed through Hormuz before the conflict had all docked by April 20th. Oil stocks will soon be at their lowest since satellite tracking began in 2018. Volumes of petrol, diesel and jet fuel at sea are already so low that gaps in supply will be inevitable. And in America petrol demand is about to surge, as the summer tempts people to get in their cars and drive.

Everybody should also acknowledge the stakes. Asia’s petrochemical industry has already idled capacity. Since the war prices of diesel and jet fuel have doubled in Asia and more than doubled in Europe. Unlike stockmarkets, where bubbles can be sustained by animal spirits alone, the price of oil is tethered to the economy at petrol pumps, docks and airports. If supply falls short of demand, prices must rise to bring about balance. There are already reports of barrels of diesel selling for $600. Good cheer cannot supplant reality.

The case for optimism is obvious. Donald Trump’s wild posting signals not just that he is rudderless, but also that he will step in whenever oil prices rise too high. Iran’s economy is broken: it urgently needs cash, which means it, too, will want a deal. If an impasse brings ruin to both sides, it will end.

The Economist is loth to second-guess those who have the facts to hand and billions of dollars at stake. However, markets have a poor record of pricing geopolitical risk. And with oil, they struggle to assess the complexities of the physical trade.

Even if a deal is in both countries’ interest, it could be hard to nail down. Each side may be underestimating the other. Mr Trump seems to think he holds all the cards. But Iran has endured long disruptions to its oil exports before, at the onset of Mr Trump’s “maximum pressure” sanctions campaign in 2018. Iran is not a democracy and the regime can survive while its people suffer. It has an incentive to hold out in the hope of a good offer for as long as it can. Mr Trump can resume the bombing, but that is as likely to delay a deal as catalyse one.

Likewise, with midterm elections looming in America, Iran’s leaders may think that Mr Trump cannot tolerate a high oil price. Yet Mr Trump is selfish. He may try to constrain price rises at home by limiting exports of refined products. The midterms are already lost, in the House at least, he may think. He is surely less bothered about the careers of Republican politicians than his own humiliation if he strikes a nuclear deal with Iran that looks worse than Barack Obama’s in 2015. His latest signal to Iran is that he is hunkering down for a long blockade.

Even if a deal is struck, the strait may not completely re-open. For one thing, the fearsome details of a nuclear pact will take months to negotiate. Now that Iran has discovered that it has leverage, it may be tempted to apply pressure with threats to close the strait again. And threats can lead to attacks. Perhaps Mr Trump will put the eradication of the nuclear programme before the complete re-opening of the strait—after all, America is an energy exporter. Supposing that America agreed to let Iran treat Hormuz as a tollgate, what then?

And even if the strait is open in principle, getting fuel into fuel tanks in practice will remain vulnerable to many unknowable delays. You can expect a rush of oil as waiting tankers escape fully laden into the Indian Ocean. But for empty tankers to return to the Persian Gulf will be more complicated. Many will have taken up bookings on other routes. The strait will need demining, which could take months. Insurance rates could be prohibitive, so governments may need to organise a scheme to cover extreme risks. Shutting down production could have damaged oil wells. Restoring output will also take time. Partially mothballed refineries won’t immediately return to full capacity.

The world is only starting to get to grips with what may lie ahead. Central banks may soon face the second inflationary shock of the decade, after the covid-19 pandemic. In Asia many governments have already taken drastic measures, such as shortening the working week. Europe’s governments will also have to change gear. So far they have focused on supporting consumer demand. They may have to deal with demand destruction—and, given the possibility of shortages of diesel and jet fuel, plan to protect food-delivery and vital services.

It will never happen, will it?

Bullish investors could be in for a nasty shock, too. The recovery from covid, Europe’s adaptation to the loss of most Russian gas and Mr Trump’s moderation of his tariffs have all led traders to trust that things always work themselves out. Amid strong corporate profits in America it may seem as if the world economy can bear any shock—and that Mr Trump will obviously back down before a catastrophe. The pain of a scenario that oil analysts have feared for decades is approaching. It will not be pretty. Get ready. ■

Michelle Goldberg - Political Violence Is Reprehensible. That Doesn’t Make Trump Less Depraved.

Michelle Goldberg - Political Violence Is Reprehensible. That Doesn’t Make Trump Less Depraved.

Read time: 4 minutes

April 27, 2026

Cole Tomas Allen, who was arrested during an attempt to storm the White House Correspondents’ Association dinner on Saturday, may be America’s first normie liberal terrorist.

The right, naturally, sees Allen as part of a pattern, lumping him in with figures like Thomas Matthew Crooks, who fired on Donald Trump in 2024, grazing his ear; Ryan Wesley Routh, who carried a semiautomatic rifle to one of Trump’s golf courses a few months later; or Tyler Robinson, charged in the killing of Charlie Kirk last year. But all those men had weird or heterogenous politics. Crooks was a nihilistic Republican misfit. Routh had a history of violence and a delusional fixation on Ukraine, where he reportedly tried to join the war effort. Robinson seems to have cooked his brain in online fetish subcultures.

But Allen, who on Monday was charged with attempting to assassinate the president, seemed to be a man with remarkably ordinary political opinions. Social media posts that appear to come from him suggest that he despised ICE, cared a lot about Ukraine, and, like the majority of Americans, wanted to see Trump impeached. Far from a radical leftist, he reposted criticisms of pro-Palestine protesters and the left-wing streamer Hasan Piker. He wasn’t exactly a standard Democrat — he was registered to no political party, and at least at one point was an evangelical Christian — but from what we know so far, before he showed up at the Washington Hilton, he had fairly mainstream beliefs.

This makes Allen’s apparent attempt at political martyrdom particularly convenient for conservatives who want to stigmatize Democratic denunciations of the president. National Review blamed “the feverish opposition to Trump” for allowing “sundry fanatics and losers to resort to political violence.” The Wall Street Journal’s editorial board tied Saturday’s attack to a political culture in which Trump’s opponents have lost “all judgment and proportion.” Some nonpartisan journalists have parroted this framing. On CNN, Dana Bash asked Representative Jamie Raskin whether he’s thinking twice about “heated rhetoric” against the president, such as calling him “terrible for this country.”

I can’t really blame Republicans for exploiting the attack; Allen has provided them with an irresistible rhetorical cudgel. The problem, of course, is that Trump is indeed terrible for this country. The fact that people have tried to kill him can’t be a reason to eschew frankness about his depravity. Rather, it’s a reason to reiterate that even depravity doesn’t justify political violence, which is morally abhorrent, socially corrosive and counterproductive.

It’s true that the manifesto attributed to Allen contains exaggerated accusations. “I am no longer willing to permit a pedophile, rapist and traitor to coat my hands with his crimes,” the manifesto declares. There is no convincing evidence that Trump has ever abused children; all the women who’ve credibly accused him of sexual assault have been adults. Calling Trump a “pedophile” has become a too-common way to describe the president’s intimate relationship with Jeffrey Epstein and coverup of the Epstein files. The manifesto is a reminder that all of us should be more precise in our language. To describe Trump accurately, however, will always sound to some like incitement.

There’s a fierce argument in America about whether the right or the left is more violent. Until very recently, there was no contest: The right was. (A 2024 study using National Institute of Justice data found that in the United States since 1990, “far-right extremists have committed far more ideologically motivated homicides than far-left or radical Islamist extremists.”) In recent years, however, there’s been an uptick in left-wing terrorist attacks and plots.

A report on this phenomenon from the Center for Strategic and International Studies suggests that there’s a ratchet at work. It noted that both Republicans and Democrats overestimate their foes’ approval of violence, and said, “Widespread polarization and misperceptions that the other side is far more violent than it actually is creates a dangerous environment where extremists can more easily rationalize using violence.” Each act of political violence further frays our threadbare social fabric, laying the foundation for authoritarianism.

After any act of political terror, conspiracists will often make “false flag” accusations, and Saturday was no different; as The New York Times reported, uses of the word “staged” soared on X. There is, of course, no defense for spreading disinformation or indulging in ideological self-delusion. Still, we can recognize that people start such rumors because they correctly intuit that violence often discredits the causes that inspire it. The left-wing terrorism rampant in the 1970s helped usher in Ronald Reagan, not socialist revolution. The 1995 Oklahoma City bombing ended up being a boon to Bill Clinton’s political fortunes. By attempting to kill Trump in 2024, Crooks helped to elect him. Violence isn’t just ethically reprehensible; it’s strategically stupid.

At least one person at the White House correspondents’ dinner was elated by the chaos the shooter caused. Describing agents running into the room with guns amid screaming and flipping tables, Dana White, the Trump-supporting head of Ultimate Fighting Championship, said it was “awesome — I literally took every minute of it in.” Perhaps he understood that the attack had given his movement a gift. Whatever evil the would-be assassin thought he was fighting, all he did was feed it.

Get used to the long Iran war

Get used to the long Iran war

Tehran has a strong incentive to keep the conflict going

By Edward Luce 

For The Financial Times

April 28

It took America 12 years from the first Gulf war in 1991 to the second to get rid of Saddam Hussein. The dictator’s capture was followed by years more of Iraqi insurgency. Starting with the financial markets, there is a lot of complacency on how soon the third Gulf war is likely to end. Donald Trump’s Operation Epic Fury on Iran has turned into an epic search for a way out. There is no obvious off-ramp, however, that does not lead back to the highway.

Iran’s latest offer illustrates why. Under the plan, Trump would lift the blockade, Iran would open the Strait of Hormuz and they would tackle the nuclear issue later on. Though Trump rejected Iran’s proposal, it is unclear at this point that he can get anything much better. It was offered shortly after the US president had for the second time had to cancel his negotiators’ trip to Islamabad. Much has been said about how unqualified Steve Witkoff and Jared Kushner are to conduct complex nuclear talks. It may be a while before we find out the extent of their inexpertise.

The war’s most salient feature is that Iran has a strong incentive to keep it going. No matter how many times Trump offers a new golden age for the Middle East, Iran will find it hard to believe he will not flip back to regime change if things do not go his way. He has been veering from promising heaven to vowing hell on an almost daily basis. That Iran’s regime is brutal and fanatical is no excuse. Even the gentlest of negotiating counterparts would find it hard to take him at his word.

From Iran’s standpoint, the longer the waterway is closed, the more likely Trump will be to digest that lesson. But let us imagine that he agrees to some variation of Iran’s recent offer. Each side would hold a threat over the other in case the nuclear talks faltered or failed. Trump’s weapon would be a resumption of bombing that would include more decapitation strikes as well as targeting bridges, power stations and even desalination plants. Iran’s weapon would be to close the Strait of Hormuz again. It does not take a grand strategist to see which threat carries more leverage. Whatever he may say about Iran’s nuclear ambitions, Trump has made it clear that reopening the waterway is his chief war aim. Iran has made it clear that it can absorb plenty of punishment from the air.  

If they do somehow manage to reach a nuclear deal, that would still leave Iran’s missile programme and regional proxies unaddressed. That Trump appears to have dropped the latter two goals as part of his framework is under-appreciated outside the Middle East. For the Gulf states, closing Iran’s missile production matters as much as giving up its stockpile of enriched uranium. Their expatriate-based economic models depend on ending Iran’s missile and drone threat. Anything less could be disastrous for business.  

For Israel, ending Iran’s support for militant groups is almost as important as killing its nuclear programme. At any point, Israel could scupper a Trump deal by fully resuming hostilities with Hizbollah. Iran would retaliate and the situation would revert to square one. Trump and Benjamin Netanyahu’s chief criticism of Barack Obama’s 2015 Iran-US nuclear deal was that it omitted those two goals. Trump is now ready to do what Obama did.

The bet is that Trump will take a temporary off-ramp in the Gulf before his China summit with Xi Jinping in mid-May. From the moment he returned to the White House in 2025, Trump has been pushing for a Xi meeting. After the two leaders agreed a trade war truce in October, Trump finally got the invitation he had been soliciting. His China state visit was originally scheduled for late March. He requested a delay to mid-May because Epic Fury was still raging. He will not want to postpone again.

Wars come to an end when one side wins or when both sides are exhausted. Short of gambling on a US ground invasion — a spectre to which Trump is rightly allergic — we can discount the chances of either side winning. Iran’s willingness to outwait Trump is thus the key question. Here again China’s role is critical. Trump’s blockade on Iranian vessels and ports in practice seems to exclude ships that are destined for China. Seizing Chinese property would be tantamount to declaring war. Trump will not risk that. Which means Iran can rely on Chinese revenues to help keep it going.

Meanwhile, Pakistan, effectively a client state of China, as well as mediator in the talks, just gave Iran permission to use its territory as a land route for trade. As has often been observed, Trump started this war but Iran will decide when it ends. 

edward.luce@ft.com