AI Has Entered the Age of Great-Power Rivalry

Every business, government, investor, and venture capitalist is chasing the same thing. Artificial intelligence has entered an age of rivalry, risk, and power — and like it or not, the world now revolves around it.

AI’s trajectory can no longer be separated from great-power rivalry, military competition, economic statecraft, and competing models of governance. If the US intends to remain the world’s foremost technology power and win the broader geostrategic contest, it must win the AI race.

There is no other option.

A new AEI analysis examines AI’s strategic implications: how it’s reshaping military operations, economic productivity, alliance structures, and deterrence — while introducing new risks of instability, escalation, and misuse. Geopolitics is shaping the technology right back: rivalries among major powers are driving research priorities, regulatory approaches, and international norms, determining whether AI becomes a source of cooperation, coercion, or conflict — and who comes out ahead.

The result is a fast-moving set of questions about regulation, tariffs, export controls, trade, sanctions, cybersecurity, AI governance, and warfare. Policymakers, analysts, and business leaders need to understand how AI is reshaping the global order. The answers are still emerging, but the broad strokes are visible.

The Stakes

AI is on track to become one of the most consequential technologies in generations, with sweeping implications for the economy and world politics. According to Moody’s, AI investment topped $581 billion last year. Global enterprise spending is projected to exceed $1.4 trillion this year — a scale with no historical precedent. McKinsey estimates the race to scale AI has triggered one of the largest infrastructure buildouts in modern history. Global spending on data centers could reach $7 trillion by 2030.

There will be winners and losers. That’s exactly why the stakes are so high.

AI is moving faster than most organizations can absorb. Every company and government wants to put it to work; the real challenge is figuring out where it creates value, how to scale it, and what has to change internally. Leading companies, consultancies, and research institutions have moved past pilots and are now installing AI across entire enterprises at breakneck speed. The cost is enormous — but so are the gains. NVIDIA, OpenAI, Anthropic, Microsoft, Moonshot, DeepSeek and many others are all jockeying for dominance.

The economic endpoint is hotly debated. Technologists closest to AI’s development largely expect it to drive massive productivity gains alongside significant net job losses and higher unemployment. Most economists, drawing on the history of past general-purpose technologies, are more circumspect — expecting productivity gains to diffuse slowly enough for labor markets to adjust, with job losses largely offset by new employment tied to the wealth AI creates.

That domestic debate ignores a global reality. Europe has largely ceded its tech future — few major companies, little venture capital. China has not. India is emerging as a serious player. The Gulf States want in on AI and data centers but aren’t originating the underlying technology. Singapore is a smaller but rising niche player, and its central bank — the Monetary Authority of Singapore — is building a genuine regulatory sandbox for AI-driven finance.

China’s AI Strategy

China — a communist dictatorship and nuclear superpower, a rivalry Xi Jinping himself has framed in terms of the “Thucydides Trap” — is competing on multiple fronts all at once.

Open-source dominance. Rather than chase closed, frontier-only models, Chinese labs have leaned into open-weight releases paired with detailed technical reports. Alibaba’s Qwen family is the standout: over 100,000 derivative models on Hugging Face — the largest open-weight ecosystem on the platform, surpassing Meta’s Llama — and more than 100 million monthly active users. Even US firms are adopting it; Airbnb uses Qwen for customer service because it offers comparable capability at lower cost.

An “open collaboration loop.” Analysts describe a compounding effect in which open models accelerate one another’s progress: frontier labs refine each other’s base models, enterprises adapt them for niche uses, and deployment data feeds back into further capability gains. A recent report identifies this as one of two loops driving China’s strategy, citing DeepSeek’s derived models as an example.

State-directed industrial policy. China’s 15th Five-Year Plan (2026–2030) embeds AI as a foundational priority alongside defense and economic growth — pushing self-reliance in algorithms, model science, and semiconductors toward an “independent and controllable” ecosystem. Officials have pledged “extraordinary measures” to lead globally, and researchers note a tonal shift from “catching up” to genuine confidence.

Working around chip restrictions. US export controls cut China off from Nvidia’s most advanced chips, though the line has moved: Washington approved H200 sales to China in a deal formalized in early 2026. Domestically, chipmaker Cambricon plans to deliver 500,000 AI accelerator units this year, though Chinese chips still lag Nvidia’s Blackwell and Rubin GPUs on key performance metrics.

Platforms over products. Chinese firms are shifting from exporting products to exporting capabilities and ecosystems. Domestic adoption is running at massive scale: smart device shipments are projected to hit 900 million units in 2026, AI endpoint penetration is expected to exceed 93% by 2027, and robotics spending is set to grow sharply.

Global governance leadership. At the World AI Conference in Shanghai, Xi paired his address with a Chair’s Statement, action plans, an AI-agent interoperability initiative, and the launch of the World AI Cooperation Organization — a coherent, China-led governance vision positioned as more open than Washington’s competition-and-security-oriented approach. Clearly, Xi’s speech was a bid to de-center the US from the global order.

India’s Rise

India has become one of the more notable AI stories — competing on adoption and talent rather than frontier model development. Stanford’s Global AI Vibrancy rankings place India third globally, behind only the US and China. The SIDE index, covering 71 countries, ranks it fourth on the standalone AI index — behind the US, China, and Singapore, but ahead of Germany, France, Japan, the UK, and Canada — and fifth in overall digital economy rankings.

Talent, not chips. India’s core strength is human capital. AI talent concentration has more than tripled, and Indian developers were the second-largest contributors to global AI projects on GitHub, at nearly 20%. NASSCOM projects India’s AI professional base will roughly double to over 1.25 million by 2027 — already the world’s second-largest AI talent pool.

Government push. The India AI Mission, backed by roughly $1.2 billion in government funding, coordinates compute infrastructure, talent pipelines, language models, and startups under a single national mandate. India’s AI market is projected to grow from about $6 billion now to $17 billion by 2028. Notably, India has opted for a risk-tiered advisory framework over mandatory pre-deployment licensing — a lighter regulatory touch than its peers.

Convening power. India recently hosted the AI Impact Summit, drawing over 35,000 registrants from more than 100 countries — positioning itself as a policy convener, not just a participant. Former UK PM Rishi Sunak has argued India is well-placed to lead, noting that leadership is about deployment as much as invention, backed by deep talent, strong digital infrastructure, and public enthusiasm that contrasts with anxious Western sentiment.

The caveat. India isn’t yet a frontier-model player on par with the US or China — it has no OpenAI, DeepMind, or DeepSeek-caliber lab — except for Purple Fabric — producing cutting-edge foundation models. Its edge lies in adoption, deployment, and talent supply — multilingual tools like the government’s Bhashini platform, fintech, agriculture, and health applications — rather than compute or foundational research. India’s long-term competitiveness hinges on mobilizing risk capital, expanding compute access, and building stronger university-startup and commercialization pathways. To the ledger of advantages India has, the fact that it has banned hundreds of Chinese apps including their AIs strikes me as an excellent idea and one the US should copy.

The Bottom Line

The global order is being reshaped by artificial intelligence.

The US-China rivalry is the main event; everyone else is playing catch-up. China is betting its status and its money on finishing first. The US cannot afford to let that happen. Political, economic, military, and technological leadership of the world all hinges on this outcome.

David Sacks, former US crypto and AI czar, has often emphasized the need for the U.S. to win the AI race by out-innovating competitors, building robust AI infrastructure, and fostering a large AI ecosystem. He believes that losing this race would significantly impact the US economy and global power dynamics.

The President’s Council of Advisors on Science and Technology (PCAST) — re-established last year and including prominent tech leaders — advises the president on science and technology, with particular focus on AI’s impact on the workforce. The question is whether that’s enough, or whether stakes this high demand a new Manhattan-level Project to coordinate AI activity across the entire private sector and government alike. We don’t want or need a Bill Gates–inspired collectivized AI structure or oversight, but do need a strategy to win. This is a no-lose proposition.

Alex Karp, the CEO of Palantir, summed it up best when he said, “We are going to be the dominant player, or China is gonna be the dominant player. And there will just be very different rules depending on who wins.”