The Trump administration has fundamentally reframed what the United States expects from its allies, making access to cutting-edge artificial intelligence the new currency of international partnership.
For generations, shared democratic values and military security concerns defined the Western alliance structure. Today, as frontier AI models become central to economic and military power, the White House is using control of those systems as a lever to reshape global relationships.
The shift is unmistakable in recent moves. The administration lifted export controls on Anthropic's Fable and Mythos models on Tuesday, part of Project Glasswing, which is set to expand access to Mythos among 150 additional organizations across more than 15 countries. Yet simultaneously, the White House is restricting European allies from the world's most powerful AI systems while demanding they conform to American geopolitical priorities.
"The problem we have is that we are leading everybody by a lot," Trump said in a recent interview with The Axios Show. "Europe has to be very careful. They're losing their way entrepreneurially." He criticized specific European policies, including the U.K.'s reluctance to develop energy resources in the North Sea due to environmental concerns, framing the region as falling behind in the technological competition that now matters most.
The tension is stark: Europe is simultaneously being asked to join Pax Silica, a U.S.-led initiative to secure AI supply chains and critical minerals, while being locked out of America's most advanced AI models. The European Commission said Wednesday it is seeking expanded access to Mythos through Project Glasswing, currently relying instead on access to GPT-5.5-Cyber. An EU delegation is now in the United States to negotiate the parameters of future technology discussions, including frontier AI models, chip supply chains, and cybersecurity.
"Our sovereign legislation is not up for negotiation," European Commission spokesperson Thomas Regnier stated, even as his bloc tries to gain closer footing with American AI companies and the Trump administration.
The strategy reflects a broader transactional recalibration of alliances. Vice President JD Vance signaled the confrontational posture toward Europe at last year's Paris AI Summit, and Trump's second term has intensified the approach. While the administration has rejected European AI safety regulations in favor of rapid innovation, it has also created its own licensing regime that generates uncertainty. OpenAI's GPT-5.6 faced a staggered rollout last month due to government concerns, showing that even American companies operate within constraints the White House sets.
Some allies are adjusting expectations. The United Arab Emirates' assistant foreign minister for advanced science and technology, Omran Sharaf, told Axios that "it's very important that trusted partners and strategic partners are included in the process," suggesting that coordinated standards around AI access would be preferable to unilateral restrictions.
For nations watching from abroad, the calculus is becoming clearer. As A.J. Bhadelia, head of global government affairs at Cohere, noted, governments are concluding they will participate where they can in the U.S. ecosystem while building independent capacity around what remains inaccessible. The alliance structure is not dissolving, but it is being reorganized around technological hierarchy rather than shared values.
Author James Rodriguez: "The White House is betting that AI dominance will keep allies aligned even as it freezes them out of its best tools, but that gamble assumes Europe won't accelerate its own models faster just to escape the dependency."
Comments