The Billionaires’ Empire of AI

The New Empire

Artificial intelligence is sold as liberation. Journalist Karen Hao has already likened today’s AI giants to the East India Company. It is not a throwaway line. It is a warning.

The East India Company began as a trading venture, a private enterprise chartered by the Crown. Within a century it ruled millions in India, extracting resources, dictating law, and fielding its own army. It was private power interlinked with state authority, a commercial machine that became sovereign.

AI companies are marching down the same path. They do not govern territory, but they govern populations — jobs, culture, thought itself. They are private empires financed with public money, commanding infrastructure so vast that governments are reduced to supplicants.

To grasp the risks of AI today, one must understand how the East India Company operated: a corporation turned empire, justified in the language of trade while consolidating political control.

Billions from Billionaires

The evidence is visible. Microsoft has poured $13 billion into OpenAI. Amazon has staked $4 billion on Anthropic. Google continues to bankroll Gemini and DeepMind. Elon Musk burns capital on xAI. Meta spends $20 billion a year to keep AI at its core. Jensen Huang of NVIDIA, wealthier than many countries, controls the GPU chokepoint that makes it all run.

Scale in Infrastructure: Manhattan-sized Data Centers

Mark Zuckerberg has announced that Meta is planning data-center clusters so large that one “covers a significant part of the footprint of Manhattan.” The facility, known internally as “Hyperion,” is intended to reach power usage on the order of several gigawatts — comparable with a small city.

And Meta is not alone. Microsoft is pouring tens of billions into hyperscale facilities. Google is redirecting infrastructure budgets to grid-linked AI super-clusters. The scale is no longer corporate. It is metropolitan.

The Money Being Spent

To give an example of the money that is about to be spent — and the money that will be amassed — consider where this capital comes from. It is investor money. It is the wealth stored in pensions. It is the total savings of countries, funnelled through equity markets into the balance sheets of these firms. Your retirement fund, your insurance premiums, your deposits — all underwriting the most resource-hungry infrastructure build in modern history.

The numbers are staggering:

• McKinsey estimates companies across the compute chain will need to invest $5.2 trillion by 2030 into data centres purely to meet AI demand.
• Microsoft alone is planning around $80 billion in 2025 for AI-enabled data centres.
• Google has earmarked $25 billion over two years for AI infrastructure tied into a major U.S. electric grid.
• Meta speaks openly of spending hundreds of billions on multiple mega-centres to fuel its “superintelligence” ambitions.

These are not abstract projections. They are the largest corporate infrastructure outlays in history — funded by collective savings and designed to lock populations into private empires of compute.

Colonising the Population

The East India Company collected taxes and rewrote laws. AI firms shape populations by subtler means: curating datasets, embedding “safety layers,” setting moderation rules. A chatbot refusal or slanted answer looks trivial, but repeated billions of times it defines culture.

Control of AI means control of discourse. Models will mediate education, medicine, employment, and law. A student’s lesson, a patient’s query, a lawyer’s draft — all filtered by corporate priorities. The limits of acceptable thought are no longer drawn by parliaments but by boards of directors.

A Lesson from Empire

The mechanics of the East India Company bear repeating. Chartered by the Crown, it traded under the guise of commerce while acquiring sovereign powers: levying taxes, raising armies, enforcing law. By the late eighteenth century it was effectively a government.

That history is precedent, not metaphor. It shows how private actors, once entrusted with infrastructure, can morph into authority itself. Today’s AI giants are repeating the pattern with data, compute, and cultural mediation.

Musk Crosses Into Politics

This evening, Musk appeared virtually at the Tommy Robinson rally from the United States, seemingly to offer encouragement to Robinson and his supporters. It was not the use of a platform. It was the use of his presence. A global industrialist stepping into a controversial political arena, signalling clearly how his stature can be leveraged.

This is the same logic of empire: wealth transmuted into authority, presence weaponised into power. What Musk has demonstrated through visibility and stature, any billionaire can embed through the technologies they command.

Musk, Zuckerberg, and Political Engineering

Elon Musk demonstrates how ownership reshapes discourse. The moment he bought Twitter — now X — the platform mirrored his worldview. Moderation rules bent, banned accounts reappeared, and his voice dominated feeds. Twitter became an extension of its owner.

The same logic applies to AI. Fine-tuning a model or adjusting its policy layer is enough to tilt tone and push ideology. What Musk did to social media, any billionaire can do to AI assistants that mediate daily knowledge.

Zuckerberg’s Facebook has faced allegations of tilting elections. Even marginal bias in information flow can alter outcomes. Now imagine AI assistants, not feeds, deciding how information is delivered. Influence becomes systemic, embedded in infrastructure rather than headlines.

Those who own the platforms decide defaults. Defaults shape culture. Culture hardens into power.

Labour and Resource Frontiers

Empire thrives on extraction. For the East India Company it was land and labour. For AI it is data and human attention. Kenyan workers absorb trauma, filtering toxic content for a few dollars an hour. Artists see their work scraped into datasets, often without consent.

The ecological cost is immense. Data centres consume electricity and water at the scale of cities. The International Energy Agency warns global demand could double by 2026. Google used five billion gallons of water in 2022 alone to cool its facilities. In places like Chile, AI firms already compete with communities for survival.

Every model query carries a hidden price, just as every East India Company ship carried stolen grain.

Public Money, Private Gain

The empire is not funded by billionaires alone. It is subsidised by states. The U.S. CHIPS Act pledges $40 billion in subsidies and $75 billion in guarantees. The EU mobilises €43 billion. The UK, Japan, and South Korea add their own packages.

Taxpayers build the backbone. Billionaires privatise the profit. It is the same colonial formula: socialised costs, private extraction.

The Stakes for Sovereignty

The East India Company existed because the British Crown allowed it, and eventually relied on it. Governments today are making the same error. They fund AI firms lavishly, then outsource real power to corporate “safety teams.” Legislatures speak of sovereignty while conceding it line by line in contracts and APIs.

Control of AI is control of sovereignty. Whoever manages the infrastructure decides which jobs remain, which narratives circulate, which questions can even be asked.