World Collapse Expert: We’re Entering The Most Dangerous Global Power Vacuum Ever

Why does it matter? Because the architect of the global order is tearing it down from the inside — and no one is ready for what comes next.

Ian Bremmer, one of the world's leading political scientists and author of the annual Top Risk report, sits down to explain why the world is more unstable than at any point in recent memory — and why the source of that instability isn't Beijing or Moscow. The picture he paints is unsettling precisely because it is structural, not cyclical.

• The United States has become the single biggest driver of geopolitical risk in the world — not China, not Russia.
• Trump went into Iran for three specific reasons, and understanding those reasons predicts what he'll do next.
• An Anthropic AI model so dangerous it couldn't be released triggered emergency meetings at the Fed and JP Morgan — and almost nobody noticed.
• China's dominance in critical minerals and green technology is already decades in the making; the dependency gap is structural, not fixable in a single policy cycle.

America isn't being challenged by China — it's dismantling its own world order

The United States built the postwar global system — the free trade architecture, the security umbrella, the open-borders immigration model. Now, Bremmer argues, no external rival is tearing it down. The Americans themselves are.

"The Americans are saying, 'We no longer want to play by the rules that we set up historically. We don't want the free trade system that we put together. We don't want to be the global policeman.'" The Chinese aren't forcing this retreat. Washington chose it.

The consequence, Bremmer says, is a shift from a world with a dominant stabilizing force to what he calls a G0 — no G7, no G20 with meaningful authority, just an absence of global leadership. "If the Americans are no longer willing to act as the global leader, but no one else is capable of filling those shoes, you don't have governments coming together and agreeing on rules of the road. You have the powerful making rules useful to them and the weak having to accept it."

Every country, economy, and security arrangement built on the assumption of American leadership is now floating on uncertain ground. Bremmer is direct: this is the most important risk of 2026, not because it might happen but because it already is. "There's no way you could look at the geopolitical order today and not say this is the most important thing — creating real movement in how the global economy works, how global politics works, global security, everything."

Trump went into Iran because Venezuela worked — and nobody around him was willing to say otherwise

There were three reasons Trump launched military action against Iran, and Bremmer lays them out with clinical precision.

First: Venezuela. Early in the term, Trump ordered the removal of Maduro. Not a single American serviceman died. Maduro ended up in a Brooklyn jail. A new government pledged cooperation on oil and mining. Leaders across South America called it a success. Trump took one lesson from this: he could replicate it at a larger scale in Tehran.

Second: Iran had already blinked, repeatedly. In Trump's first term, they threatened devastation after the assassination of Qasem Soleimani — and did nothing. During the 12-day Israeli campaign last June, they fired missiles at the Al Udeid base in Qatar and warned the Americans in advance through Iraq. "It was very clear the Iranians didn't want any part of that fight." So Trump told himself, "This is going to be awesome. I'm going to go in, I'm going to pull Venezuela in Iran. And I know they don't want to fight me."

Third, and most consequential: the cabinet. In his first term, Bremmer notes, figures like Pompeo and Mattis were "patriotic first and foremost" — they pushed back, leaked, and occasionally undermined decisions they believed would harm the country. This cabinet is different. "What they all share is that they are first and foremost loyal to the president." What Trump hears is shaded toward how brilliant he is. "He hears very little of that and he's taking away, 'I'm incredible. I'm confident I'm going to make this happen.'" The most dangerous leader isn't the most aggressive one — it's the one whose feedback loops have been captured.

Anthropic built a model so dangerous it couldn't be released — and central banks are already treating it as an emergency

Jerome Powell and Scott Bessent called an urgent meeting of every major bank CEO. Jamie Dimon — who runs the institution Bremmer describes as "by far the best at cybersecurity among the big banks" — called it a five-alarm fire. The trigger: an Anthropic AI model capable of identifying exploitable vulnerabilities across every piece of software in existence.

"They created a model which is so powerful that they couldn't release it because it would have been an immediate systemic risk to the global economy and our security."

Bremmer acknowledges the marketing dimension — Anthropic had just been removed from a Defense Department contract for refusing to let its AI be used in targeting and surveillance, so the timing was convenient. But he doesn't let that observation dilute the core threat. "This risk is real. If suddenly your systems are hackable by anyone that has access to this tool, your markets are going to go down. Your banks aren't going to work. Your data is going to be stolen."

The model needs to be deployed defensively — to find and patch the bugs before adversaries develop equivalent tools. "Other people will have these tools in very short order." The Fed and Treasury already know this. The headlines are still on Iran.

China's long game is already mostly won — and the West only notices when the supply chain breaks

China isn't ahead on every dimension. Its military has never fought a naval war. It lacks a convertible currency. Its domestic dissent is suppressed rather than resolved. Bremmer is clear-eyed about all of this. But on the question that will determine the next fifty years, the verdict is already in.

"The Chinese are now either at parity or ahead of the Americans and everyone else by a long margin in many of the core technologies that matter most in the world." Electric vehicles. Batteries. Solar. Critical minerals and rare earths — not just extracted, but reprocessed, controlled, and gatekept.

The comparison Bremmer reaches for is visceral: Europe's dependence on Russian gas. They knew the dependency existed. They didn't fix it. When Russia invaded Ukraine, the vulnerability became catastrophic. "If they determine that they're going to shut you off, you're dead."

The West's version of this story is playing out across lithium, antimony, and every material that goes into advanced weaponry, EV batteries, and consumer electronics. "The Chinese have been investing at scale globally in that capability for decades now, thinking long term. And a lot of the rest of us have not been thinking long term." Quarterly returns versus generational strategy — that asymmetry, compounded over thirty years, is now structural. Diversifying supply chains isn't forward planning anymore. It is damage control.

Trump will probably exit Iran by calling it a win — while quietly handing Iran a toll booth over the world's most important oil chokepoint

The ceasefire is holding. The 21-hour talks in Pakistan were substantive — "if these talks were a disaster, they don't last 24 hours." Bremmer's most likely scenario: the Iranians compromise on nuclear enrichment in exchange for formalized, privileged control over transit through the Strait of Hormuz.

"I expect that the Iranians are more likely to give on the nuclear issue and on enrichment if they're able to maintain a privileged position on transit through the strait because that will help provide them with money and with security."

Trump has already laid the rhetorical groundwork for retreat. He's said the nukes are buried under rubble and no longer a priority. He's said regime change has already happened and there are new people to work with. He's said the strait isn't his problem. "He's already set the stage to back out."

The economic pressure is real and entirely self-inflicted. "This is an economic downturn with oil prices shooting up, gas is over four bucks a gallon, diesel's over five, inflation's ticking up — he is completely responsible for it. No one else is responsible." Unlike the pandemic, there is no one else to blame. Iran knows it, watches his polling numbers, and calculates that they don't need to wait long. The deal that eventually emerges will be framed as a Trump triumph. The structural reality will be Iran with a durable revenue stream and leverage over a chokepoint that moves a significant fraction of the world's oil.

AI is programming its users — and that is a more immediate threat than AGI

Bremmer's sharpest concern isn't superintelligence. It's something quieter and already underway.

"I worry the most about people getting programmed. I'm not worried about artificial general intelligence. I'm worried about human beings becoming more computer-like. When you spend all of your time on your smartphone, that is a computer programming a human being."

The algorithm doesn't want you to change your views. It rewards certainty, punishes ambiguity, and filters out friction. The result, Bremmer argues, is a homogenization of thought that democracies and economies depend on to self-correct. "As soon as it gets intermediated by algorithms, as soon as you get programmed into a lane, we become much more inhuman."

He singles out prediction markets as a symptom — political institutions reduced to a casino where the only question is whether you're in or out of the money. This is what he means when he says, toward the end of the conversation, that the digital world is not really a human world. The antidote, as he and Steven discuss, is deliberate: long-form conversation, voluntary exposure to disagreement, seeking out people whose conclusions differ from yours. Not because it's pleasant. Because the alternative is becoming an output of the machine rather than a person using one.

The next populist earthquake hits in 2028 — and it will be driven by people no current politician knows how to reach

Trump's coalition was built on men without advanced degrees who watched automation and free trade hollow out their economic futures. That wave has already broken. The next one is forming.

"We haven't seen women with advanced degrees, urban and suburban, worried about their jobs and worried about their kids. And that wave of populism is coming absolutely in 2028."

A pro-technology, pro-business, centrist US senator told Bremmer recently that they can no longer discuss data centers publicly. "I've never seen my constituents so upset about an issue." No jobs created locally. Energy prices rising. Water usage spiking. Zoning disrupted. AI is now less popular than ICE in the United States.

When AI-driven displacement reaches knowledge workers and urban professionals — the people who currently assume they are on the safe side of the automation divide — the political consequences will dwarf what 2016 produced. "I don't yet know who the political figures are that are going to respond to that. I don't think that person today exists in the political spectrum."

The shape of the backlash is visible. The leader who channels it is not.

The G0 world doesn't wait for a replacement leader to emerge

What this episode reveals, underneath all the specific crises, is that the transition from American-led order to something else will not be orderly. There is no heir apparent — not China, not Europe, not a reformed multilateral institution. The gap between the system that's ending and whatever replaces it will be filled by power, not governance. Every country, company, and individual making long-range decisions right now is doing so on a foundation that has quietly been removed. The most important question of the next decade isn't who leads the world. It's how long the absence goes unaddressed before it becomes irreversible.

---

Topics: geopolitics, US foreign policy, Iran conflict, China strategy, AI risk, critical minerals, populism, Trump, global power vacuum, technology governance