With the world fixated on Iran, the Strait of Hormuz, and Donald Trump’s peripatetic presidency, it’s worth a reminder that the geopolitics of artificial intelligence (AI) will almost certainly have far more lasting consequences than any of that.

We are living through a transformation of similar magnitude to the Second Industrial Revolution of the late nineteenth and early twentieth centuries. That said, the AI Revolution is moving far faster with far less time for global leaders to adapt, so the perils are greater—notwithstanding AI’s considerable promise.

The Industrial Revolution rewired history. It shifted power from agriculture to industry, from landed aristocracies to barons of manufacturing, and from empires to nation-states that could produce, mobilize, and innovate at scale. Out of that upheaval rose capitalism and communism, mass politics and mass warfare, and eventually the cataclysms of two world wars.

So what does the geopolitics of the AI Revolution look like? The twentieth century’s race for energy, raw materials, and nuclear capability will only intensify, given data centers’ voracious appetite for power. But it will be accompanied by a contest for computational power, data dominance, and algorithmic superiority. Nations that lead in AI will not only grow faster economically, but they will also govern differently, fight differently, and project influence in ways that are difficult to predict.

It’s in that context that one should reflect on the recent emergence of Mythos, Anthropic’s frontier AI system whose capabilities are so potentially disruptive and powerful that its creators are currently restricting who can use it. According to reports, it can autonomously identify and exploit previously unknown software vulnerabilities—thousands of them—across systems that underpin the basic functions and operations for global finance, communications, and critical infrastructure. That means that your banks, internet services, and water treatment systems all grow more vulnerable.

At the Atlantic Council during the International Monetary Fund–World Bank Spring Meetings earlier this month, global financial leaders and central bankers raised concerns about Mythos in private conversations as often as they did slow growth, energy prices, and international debt levels.

Mythos provides a glimpse into a future world in which the offense-defense balance in cyberspace, military battlefields, and beyond tilts dramatically toward those who can harness and exploit AI capabilities first.

That is why it strikes me that Mythos, and not Iran, is the most consequential geopolitical warning signal for this moment. (Though it’s worth noting that a senior US official told me that OpenAI and Google aren’t far behind with similar capabilities.)

Wars like the one that has been unfolding around Iran are deadly, dangerous, and significant. But they are also familiar. We can analyze their dynamics, calculate what constitutes deterrence, and thus limit their escalation and avoid miscalculation by working from a well-known playbook.

Don’t believe for a moment anyone who wants to reassure you that we are prepared for the new world of challenges AI is introducing. Democratically elected leaders don’t comprehend the systems they are tasked with regulating. The unfolding court case between tech entrepreneurs Elon Musk and Sam Altman in Oakland, California, this week, underscores how petty competition among the giants of the AI industry masks our ability to discover some sort of national common cause. Distrust is rising among Americans about AI, and Gallup polling shows that almost half of Gen Z workers believe the risks of AI in the workforce outweigh the benefits. Authoritarian regimes, and particularly China, will have less trouble with public opinion and are moving quickly to leverage AI to gain strategic advantage against the United States—and to surveil and control their own people. And while US AI companies might have the commercial edge when it comes to large language models, China is moving ahead when it comes to applying AI to robotics and manufacturing—and potentially warfare.

As a preternatural optimist, I’m encouraged by how AI can super-empower the world’s do-gooders. But the lingering pessimist in me is equally concerned by what super-empowered evildoers might unleash. 

The promise is as breathtaking as the peril. What makes this moment different is the magnitude of the good and bad that could result, and the reality that both could arrive simultaneously.

AI is already accelerating the following breakthroughs: earlier detection and more personalized treatment of cancer, new pathways to slowing and solving Alzheimer’s, the design of life-saving drugs in months instead of years, and the optimization of everything from energy grids to food production. It could expand creativity in ways we are only beginning to imagine and unlock productivity gains that lift living standards globally.

At the same time, AI poses the following generational risks: the weaponization of biology, the erosion of truth amid synthetic media, the concentration of power in the hands of a few countries or companies, and the displacement of workers that outstrips our ability to adapt.

Here’s the inflection point: Do the United States and its partners treat AI as a competitive race to be won at any cost, or as a transformative force that requires new rules and guardrails, new forms of international cooperation, and even new institutions?

History offers a warning. The Industrial Revolution produced extraordinary wealth and innovation, but it also spawned ideological extremism and conflict on a scale the world had never seen, including Hitler’s industrialized death camps and the war-ending attacks on Hiroshima and Nagasaki.  

It took decades and two world wars before the United States and its wartime allies ushered in a stable international order, which included the founding of the United Nations, NATO, the Bretton Woods institutions, and other organizations. The AI era may not provide us with that much time.

The existing global system of rules and institutions is already strained by mistrust, fragmentation, and now war. It lacks the mechanisms to manage a technological challenge of such epochal scope. US markets are rewarding speed and scale but not international coordination.

If we face this transformational moment with transactional instincts, we risk repeating the darker chapters of the twentieth century—this time at digital speed. 

Frederick Kempe is president and chief executive officer of the Atlantic Council. You can follow him on X @FredKempe.

This edition is part of Frederick Kempe’s Inflection Points newsletter, a column of dispatches from a world in transition.

Further reading

Image: CEO and Co-Founder of Anthropic Dario Amodei speaks during the World Economic Forum meeting in Davos, Switzerland, on January 20, 2026. (REUTERS/Denis Balibouse/File Photo)