Why Letting the US and China Control AI Alone Could End Everything

Jan de Vries · 2026-09-15

Imagine a technology so powerful it could either solve our biggest problems or wipe us out. Now imagine only two countries deciding how it's used. That's the reality we're facing with AI governance today. Vendan Ananda Kumararajah recently made a compelling case that this approach is a recipe for disaster. And honestly? He's got a point. ### The Problem with a Two-Horse Race When the US and China dominate AI development, we get a sprint, not a marathon. Both nations are racing to outdo each other, often prioritizing speed over safety. That's like letting two teenagers drag-race sports cars without any traffic laws. Sure, it's thrilling, but someone's going to get hurt. Kumararajah argues that this dynamic squeezes out other perspectives. Europe, Africa, South America—they're all left watching from the sidelines. But AI doesn't care about borders. A rogue algorithm doesn't check your passport before causing havoc. So why should only two countries have a say in how it's built and deployed? ### Diversity Isn't Just Nice—It's Necessary We've all seen what happens when a small group makes decisions for everyone. Think about the early days of social media. A handful of companies in California decided how billions of people would communicate. The result? Misinformation, polarization, and a whole lot of regret. AI is that times a thousand. If we let the US and China set all the rules, we're inviting the same mistakes on a global scale. Different cultures have different values. What's acceptable in one place might be appalling in another. We need those voices in the room. > "AI governance isn't a tech problem. It's a human problem. And humans are everywhere." That quote from Kumararajah sums it up perfectly. We can't treat AI like a private club for superpowers. ### The Catastrophic Risks We're Ignoring Let's talk about the elephant in the room: existential risk. Some experts believe advanced AI could pose a threat to humanity itself. Whether it's autonomous weapons or systems that outsmart us, the stakes couldn't be higher. Kumararajah points out that without broad international cooperation, we're basically playing Russian roulette with the future. A single misstep by one country could have ripple effects worldwide. And if those countries are locked in a zero-sum game, they're less likely to share safety research or coordinate on safeguards. Here's what we need instead: - **Global participation**: Every region should have a seat at the table. Not just the big players. - **Diverse ideas**: Different cultural and ethical frameworks lead to better, more robust solutions. - **Catastrophic risk safeguards**: We need binding agreements, not just handshake deals. ### What Can We Do About It? It's easy to feel powerless when you're not a world leader. But there are things we can all do. Support organizations pushing for international AI treaties. Demand transparency from tech companies. And most importantly, keep talking about this. The more people who understand the stakes, the harder it is for politicians to ignore. Kumararajah's warning isn't just doom and gloom. It's a call to action. We still have time to get this right. But we can't afford to wait until it's too late. So next time you hear about AI regulation, don't tune out. This isn't just about tech policy. It's about our future. And that future should be decided by all of us, not just two countries.