Musk’s USAID Cuts May Hand China The AI Race

How do you win the global AI race? If you asked Elon Musk or his supporters in the Trump administration, they’d probably talk about deregulation, or getting government out of the way, or maybe something about GPU clusters and massive power consumption. What they probably wouldn’t mention is USAID, the US government’s primary foreign aid agency. And yet, it turns out that killing USAID might be one of the most effective ways to lose the AI race to China.

Many people in the AI world believed that having Donald Trump in the White House, along with Elon Musk (who owns an AI company) and surrounded by other supporters who are big on AI, like Marc Andreessen and David Sacks (who has been appointed AI czar, despite having little experience with AI), would mean a huge boost for US AI.

The irony here is rich: while Musk’s xAI and other American AI companies are desperately seeking new markets and training data, his crusade against USAID is systematically dismantling one of America’s most powerful tools for securing both. As Kat Duffy explains in a compelling Foreign Policy essay, killing USAID may have just kneecapped America’s AI ambitions. The connection? USAID has been America’s secret weapon in opening up and maintaining crucial global markets:

The United States’ untapped superpower in the AI race is not the so-called innovation economy, or an unregulated market. Those have been leveraged. Rather, it is the sheer scale of the U.S. government’s global presence, and how that could be used to turbocharge America’s global leadership in the AI race.

Moreover, the ability to build markets—not just individual AI models—will determine which nation’s technology dominates globally. Supporting the early and successful deployment of basic, useful AI systems across global markets will be key to protecting market entry opportunities for American AI technologies for decades to come.

The United States’ sturdy global presence provides a ready-made distribution network for AI innovation and adoption, particularly in critical sectors such as global health, food security, and climate resilience—areas where AI can drive immediate and transformative change.

This isn’t just about general market access — it’s specifically critical for AI development:

Broad reach also ensures that the United States’ scale is being used to benefit broader U.S. private sector interests. AI incumbents and Big Tech can hire local offices, lawyers, and lobbyists to support their expansion into new markets (and the new sources of data they offer for model training). Key benefits of the U.S. government’s scale accrue to “Little Tech.” The country’s start-ups and innovators are better equipped to balance the scales when taxpayer investments do the heavy lifting of building awareness of U.S. business, developing trust with local governments, and supporting broad private-public partnerships that facilitate rapid scaling.

While U.S. foreign assistance was not designed to support the expansion of these emerging technologies, it will be far faster to optimize for such expansion over this existing infrastructure. Financing structures, legal registrations, banking agreements, compliance protocols, training programs, public-private partnerships, growth strategies, and expert staff: These mechanics are core to foreign assistance implementation, and they could be adapted with relative speed to support AI deployment across markets. Notably, 11 of the United States’ top 15 export markets have been the recipients of U.S. foreign assistance funding to support their development.

While Musk is busy dismantling USAID in the name of efficiency, China is eagerly moving to fill the void. And not because they’re feeling particularly charitable. No, they’ve figured out something that seems to have escaped Musk’s first-principles thinking: development aid is actually a pretty clever way to build a global tech advantage.

Chinese President Xi Jinping understands that assignment. For more than a decade, the Chinese government has demonstrated how deeply it appreciates the competitive advantage that global presence confers in the digital age. Between 2013 and 2022, China invested $679 billion in infrastructure projects through its Belt and Road Initiative across nearly 150 countries. These investments have not only bolstered China’s global influence but also created dependencies with long-term strategic implications.

In particular, China’s Digital Silk Road initiative has invested billions in 5G, fiber-optic cables, and data centers across Africa, Latin America, and Southeast Asia. Those investments, notably, have also created unsustainable debt in many recipient nations, giving China greater influence over local decisions.

As a result, many governments around the world are less than eager to lean on Chinese development support. But in a vacuum of U.S. offerings—and with limited domestic capacity to grow absent foreign support—they may see no alternative.

The supreme irony here is that Musk — who has admitted to getting things wrong while continuing to spread completely false claims about USAID — is actively undermining his own AI company’s future market opportunities. It’s a masterclass in the dangers of oversimplified thinking masquerading as strategic vision.

You see this pattern a lot with the Musk and DOGE crew. They love talking about “first principles thinking,” which sounds very sophisticated and scientific. But here’s the thing about first principles: they’re only useful if you actually understand the system you’re analyzing. What we’re getting instead is what you might call “first impressions thinking” — taking whatever pops into your head after scrolling ExTwitter for five minutes and declaring it a fundamental truth of the universe. What we’re seeing is the dangerous oversimplification of intricate global relationships, amplified by an echo chamber of ExTwitter 4chan-brained yes-men who mistake contrarianism for insight.

Here’s the thing about winning technology races: Sometimes the winning strategy isn’t actually about the technology at all. China seems to understand this. They’re out there building relationships, creating dependencies, establishing presence — all the boring infrastructural stuff that doesn’t get you viral tweets or machine learning breakthroughs, but can get you actual market access and, eventually, dominance.

Meanwhile, Musk and his allies are playing a different game entirely. They’ve convinced themselves that government programs like USAID are just inefficient bureaucracy getting in the way of American innovation. It’s a view that plays well on ExTwitter, where everything can be reduced to a spicy meme about government waste and crying woke libs. But in the real world — you know, the one where actual AI companies need access to actual markets and actual data — it turns out that having a global network of diplomatic and economic relationships is kind of important.

The dumbest part is that this isn’t even a particularly complicated insight. It’s the sort of thing you might figure out by, say, looking at how America became a technology superpower in the first place. But that would require actually studying history instead of just retweeting ChudLord69. And it would require admitting that sometimes government programs actually serve a purpose beyond providing material for outrage posts.

So here we are, watching in real time as America’s AI ambitions get kneecapped by the very people claiming to champion them. It’s like a masterclass in unintended consequences, except the students are too busy tweeting “government bad” memes to notice they’re failing the course.

The tragedy isn’t just that we’re probably going to lose this particular race. It’s that by the time Musk and his allies figure out what went wrong, they’ll have already dismantled the very infrastructure they needed to win it. But hey, at least they’ll have their memes and tweets.