As we enter the second half of the 2020s, AI capabilities are increasingly becoming the key determinant of economic and military might. Hence, after years of ratcheting up US controls on exports of advanced semiconductors to China, the Biden administration, as one of its final acts in office, issued an “interim final rule” to establish a Framework for Artificial Intelligence Diffusion.If this AI Diffusion Rule (as it is known) remains intact, the US inputs needed to develop frontier AI models will be accessible only to a tight circle of allies.In fact, even most Nato and EU members would be sidelined from AI development – a prospect that ultimately could undermine America’s own strategic objectives.For example, Central and Eastern Europe – Europe’s industrial powerhouse – needs access to frontier AI chips to remain economically competitive and militarily secure over the next decade. Yet if America is no longer a reliable partner and supplier, the region’s economies could gravitate toward China.Specifically, the AI Diffusion Rule creates a three-tiered hierarchy of access. The bottom tier comprises China, Russia, and the other usual suspects, which are effectively barred from the market. The most privileged tier comprises 18 trusted allies and trade partners, including the other members of the so-called Five Eyes – America’s intelligence-sharing alliance with Australia, Canada, New Zealand, and the United Kingdom – and Asian technology powerhouses such as Taiwan and South Korea.That leaves the middle tier, which includes most Nato and EU members. Despite being close US allies, they face rigorous restrictions, including a cap on imports of high-end AI chips (50,000 per country through 2027).Moreover, the rule extends beyond hardware to include things like model weights (which embody an AI’s learned capabilities). Any country that wants to access such assets must pass through stringent, burdensome security protocols.Without these hurdles, Central Europe would be well placed to build on its robust industrial base to drive technological progress for all of Europe. Poland, among the best capitalised economies in the region, recently launched a $240mn initiative to help develop the first Polish large language model. Just days before the US tightened its AI chip restrictions, the Polish government announced plans to invest nearly $1.2bn this year in AI-based technologies. And in February, Polish Prime Minister Donald Tusk signed a memorandum of understanding with Alphabet (Google) CEO Sundar Pichai to speed up AI adoption in cybersecurity, health, energy, and other sectors.By relegating Central and Eastern Europe to second-tier status, the US has threatened the region’s economic future in at least three ways. First, the restrictions will make it more difficult to upgrade legacy industries such as auto manufacturing at a time when Europeans need to press ahead with autonomous and self-driving vehicles, AI-augmented manufacturing, and predictive maintenance technologies.Second, now that the region’s AI aspirations depend on the whims of policymakers in Washington, long-term planning and many new investments are on hold. As import-dependent US companies are also recognising, there is nothing worse for business than persistent policy uncertainty.Third, with high-performance chips out of reach, Central and Eastern Europe’s budding AI startup ecosystem may be forced to migrate to tier-one countries, crippling a strategic sector before it has the chance to mature. If its startups leave or wither away, the region will be set back at least a decade. No wonder public discontent has rippled across European capitals, from Brussels and Prague to Riga and Warsaw.Of course, US policymakers would say that they are making a calculated trade-off: safeguarding America’s technological edge and national security interests, even if that means causing friction with long-standing allies. But the AI chip restrictions also risk fracturing Nato’s cohesion and stalling its military modernisation.The war in Ukraine has shown that combat increasingly revolves around AI capabilities – from cyber defences to drones. Given that the Trump administration wants Nato’s European members to step up and do more, it makes no sense to maintain limits on Central and Eastern Europe’s ability to develop these technologies.True, most countries in the region cannot yet reach the quota imposed by the AI Diffusion Rule anyway, owing to a lack of computing power and other requirements. But as their capacities grow, they may begin to look for other partners. In 2023, Hungary alone absorbed 44% of Chinese foreign direct investment in Europe – eclipsing France, Germany, and Britain combined. Similarly, despite Poland’s strong Western orientation, China is already one of its largest trading partners, and the broader Central and Eastern European automotive cluster has robust indirect supply links to China.Given that these ties could facilitate a broader geopolitical realignment, the US would do well to devise a more targeted approach. Ultimately, AI leadership is not just about controlling the hardware and the software. It is also about setting global standards for how the technology is developed, deployed, and governed.By sidelining allies from the AI frontier, the US risks losing influence over the technology’s trajectory. That cannot possibly be what America wants. - Project SyndicateSona Muzikarova, a non-resident senior fellow at the Atlantic Council and a Mason fellow at Harvard Kennedy School, is a former economist at the European Central Bank, a former diplomat at the OECD, and a former senior adviser to the Deputy Minister of Foreign Affairs of the Slovak Republic.