From farms in Marshall, to the Mayo Clinic in Rochester, to reservation systems for the North Shore and other tourism hot spots, and all points in between and beyond, nearly every entity in Minnesota is dependent on and vulnerable to integrated internet systems, which makes concern over AI governance not just a geopolitical abstract but an unavoidable local actuality. Which is just one reason why Trump may be more amenable to cooperation with China despite the breakneck AI race between the two superpowers.
But history, not headlines, can better contextualize the event. Including from an ancient Greek historian, Thucydides, who chronicled the conflict between Athens, a rising power, and Sparta, a ruling power.
“It was the rise of Athens and the fear that this installed that made war inevitable,” the Hellenic historian stated. And indeed, in a concept coined “Thucydides Trap” by Harvard scholar Graham Allison, a similar scenario of a rising power threatening a ruling power has played out 16 times across history with 12 ending in war.
Regarding China and America, “it’s impossible not to hear the echoes of Thucydides,” Allison told me when he visited the Humphrey School during the first Trump administration.
Now, another Greek word worrying world leaders could be key to the summit: Mythos, the name AI firm Anthropic gave its newest model, which is sending shudders worldwide over its powerful potential — and peril, perhaps, if used by unfriendly nations or nonstate actors like terrorist groups.
Accordingly, it’s increasingly important that AI governance, or at least some guidelines, be discussed by the U.S. and China, the world’s two most-advanced AI powers.
It’s why 15 former heads of state, Nobel laureates or other notable global leaders recently called for “governments worldwide to manage artificial intelligence with an urgency that reflects both scientific evidence and public concern.”
As “the scale of AI capability accelerates exponentially, the current gap in governance is becoming a crisis,” the leaders wrote in an open letter.
Accompanying Trump are 17 CEOs of indispensable industries like agriculture (represented by Brian Sikes of Minnesota-based Cargill) as well as several from the tech sector. Jensen Huang of Nvidia, whose cutting-edge computer chips help fuel the AI frenzy, was initially overlooked, but was a late addition.
Artificial intelligence itself shouldn’t be overlooked, however.
“AI is one of the most underappreciated elements of President Trump’s trip to China, and with the benefit of hindsight, might prove to be one of the most consequential aspects of the visit,” Ryan Hass, director of the Brookings Institution’s China Center, said in an email interview.
Consequential, indeed, because Mythos, belying its definition, is all too real.
It could “crack the whole cyber-risk world open,” warned the governor of the Bank of England.
“What Mythos has succeeded in doing is to scale this process up so it can scan software systems and find many, many vulnerabilities,” said Shashi Shekhar, a professor in the University of Minnesota’s Department of Computer Science and Engineering. “It is a major step forward.”
Mythos “marks a turning point,” concurred Kyle Chan, Hass’ colleague at Brookings whose expertise includes China’s technology development and industrial policy as well as U.S.-China relations. “It really makes national security folks concerned about cybersecurity risks and being able to basically be this sort of superhuman hacker and find security vulnerabilities in a lot of the fundamental infrastructure for our digital world.”
Chan cited cyberattacks that could “go potentially beyond personal-data security and privacy issues; it would go down to the level of the foundation of our financial systems, our banking system, the software that might connect to actual physical infrastructure, potentially like power plants and the grid; it could compromise the government systems.”
“The rate of progress in AI capabilities, particularly after Mythos, seems to have recalibrated the Trump administration’s approach to AI,” said Hass. “They have gone from a hands-off, laissez-faire approach to a more active effort to oversee the development of AI and its impacts. This of course spills over into the U.S.-China relationship. Even if the U.S. oversees AI developments meticulously, those efforts could prove for naught unless China acts in a coordinated and parallel manner.”
Trump, continued Hass, “has tried furtively throughout his tenure as president to launch arms control talks with China around nuclear weapons. Those proposals have not gone anywhere. His reported plans to propose U.S.-China coordination on AI safety just might go somewhere.”
That would be a welcome development, especially since the perils of AI can go anywhere. “In fact, if anything, it may be a set of regional banks that would perhaps be more at risk given that they might have less investment in their overall cyber defenses than maybe some of the big Wall Street banks,” said Chan, who added regional power grids and other energy entities as susceptible to cyberattack.
“A lot of these systems depend on a whole patchwork of open-source major platforms,” said Chan. “So even at the local level, something that hits Microsoft, for example, can affect a farm in rural Minnesota.”
So even though other priorities for each country will be the front-page focus of much of the dialogue, emphasizing AI, Chan said, is important. “This is not just a geopolitical chess game between Washington and Beijing. This is really about trying to make sure, in a very interconnected world, that we try to reduce some of these risks now with these AI systems.”
And the key to reducing risks is to talk now, stressed Hass.
“There predictably will be skepticism about the usefulness of establishing another dialogue with China, given the uneven record of progress made in past bilateral dialogues,” he said. “I encourage people to place this channel in a separate category. If the effects of AI are any order of magnitude as disruptive as many anticipate, it will be important for U.S. and Chinese officials to begin comparing notes sooner than later about how to manage risks.”
Hass concluded with wisdom that would resonate anywhere, anytime — from ancient Greece, greater Minnesota and beyond, including Beijing this week.
“A key lesson of diplomacy,” he said, “is that it is dangerous to do a cold call to your counterpart in a crisis.”
© 2026 The Minnesota Star Tribune. Distributed by Tribune Content Agency, LLC.
