Two Superpowers, Two Philosophies

I’ve covered technology transitions for four decades, and I can’t recall a moment quite like this past weekend. While researchers debate whether artificial intelligence poses an existential risk to humanity, the leaders of the world’s two largest economies stepped forward with visions for AI so fundamentally opposed that they crystallize the central tension of this technological era.

The contrast could hardly be starker.

President Trump , speaking Saturday, waved off warnings from Anthropic, OpenAI, and Google—companies whose own scientists have urged caution about the pace of development. His message was unambiguous: American leadership in AI is not negotiable.

“We’re leading China in AI … and frankly, I want to keep it that way, because whoever wins AI wins,” he said.

This is a zero-sum framing, straight out of the Cold War playbook, but applied to a technology that behaves nothing like a missile stockpile.

The next day, in India, China’s Xi Jinping offered the opposite vision. Addressing Global South leaders at the BRICS Summit , he called for open-source development, international collaboration, and a shared governance framework for AI, though he was notably light on specifics. It was a message calibrated for nations wary of being locked out of the AI era by the two powers that currently dominate frontier AI development.

I’ve sat through enough industry cycles to recognize a familiar pattern. In the early days of the PC, and later the internet, we saw similar tensions between proprietary control and open standards. But those debates played out largely within markets. What we’re watching now is that same tension playing out geopolitically, with two nation-states essentially proposing different operating systems for how the rest of the world should relate to AI.

Here’s what concerns me most as someone who has spent a career watching technology diffuse globally: neither vision, on its own, is sufficient.

Trump’s competitive argument has some merit. American AI companies such as Anthropic, OpenAI, and Google have built commanding leads in frontier-model capability. There is legitimate strategic logic in not ceding that ground, particularly given how AI increasingly underpins economic productivity, military capability, and information ecosystems.

But a purely competitive posture, taken to the extreme, risks fragmenting the safety and governance conversations that need to happen globally. AI’s risks—whether around labor displacement, misinformation, or more speculative existential concerns—do not respect borders. A “we win, you lose” framework makes it harder to build the kind of international guardrails this technology may eventually require.

Xi’s appeal for openness and cooperation, meanwhile, sounds appealing on its face, especially to nations understandably anxious about being permanently dependent on either Silicon Valley or Beijing for their AI infrastructure. But China’s own domestic AI governance is hardly a model of openness, and a proposed international framework without specifics is more diplomatic positioning than actionable policy. I’d want to see real substance before reading too much into it.

Neither proposal can be evaluated solely on its stated ideals, because both are shaped by a deep strategic rivalry. What strikes me most is the trust deficit sitting underneath both speeches. Neither government believes the other is operating in good faith, and that suspicion is now shaping how each presents AI to the rest of the world. One as a race to be won, the other as a coalition to be built against the frontrunners.

If there’s a lesson from prior technology cycles, it’s that the winners over the long run tend to be the ecosystems that balance competitive innovation with enough openness to earn broad adoption. Neither Washington, D.C. nor Beijing has fully cracked that balance yet. Until one of them does, or until they find a way to genuinely cooperate on the parts of AI governance that require it, the rest of the world will be left navigating two competing approaches without a clear roadmap for either.

That is a precarious foundation on which to build an era-defining technology.