Technofeudalism, Explained: Why You Work for Big Tech for Free — and Who Really Owns AI

If you have ever posted for free on a platform that sold your attention to advertisers, or paid a monthly fee to store your own files on someone else’s servers, you have already lived inside technofeudalism. The word is deliberately provocative. It argues that the digital economy has quietly stopped behaving like a market of buyers and sellers and started behaving like something older and stranger — a system of lords, land, and rent. To understand why a growing number of serious thinkers reach for a medieval metaphor to describe our very modern lives, it helps to start with a single, clean idea.

From land to data: the one-line definition of technofeudalism

Here is the technofeudalism meaning in one sentence: it is the claim that the most powerful tech companies no longer make most of their money by selling goods in a competitive market, but by owning the digital territory the rest of us must pass through — and charging rent for access to it.

The term was popularised by the economist and former Greek finance minister Yanis Varoufakis, most fully in his 2023 book Technofeudalism: What Killed Capitalism. His argument is bold: capitalism, he says, has been displaced from within by something new. The related French economist Cédric Durand reaches a similar diagnosis from a different route — that the platforms have become a kind of infrastructure we cannot opt out of, and that owning infrastructure is a fundamentally different game from competing in a market.

Varoufakis calls the thing these companies own cloud capital: the vast stack of servers, platforms, algorithms, app stores, and recommendation engines that increasingly mediate what we see, buy, read, and do. In the old feudal order, whoever owned the land collected rent from everyone who worked it. In the new order, whoever owns the cloud collects a toll from everyone — merchants, creators, and ordinary users alike — who must cross their digital ground to reach anyone else.

The distinction he draws is subtle but important. Ordinary capital — a factory, a fleet of trucks, a warehouse — makes money by producing and selling. Cloud capital, in his telling, does something different: it positions itself between buyers and sellers and charges for the connection. An app store that takes a share of every transaction is not competing in the market so much as it is hosting the market and taxing everyone inside it. That, Varoufakis argues, is closer to a landlord’s rent than a merchant’s profit. Whether that difference is large enough to justify a whole new name is the question the rest of this piece will keep circling back to.

Whoever owned the land collected rent from everyone who worked it. Whoever owns the cloud collects a toll from everyone who must cross it.

How you became a digital tenant

The uncomfortable heart of the idea is that you are not really the customer. You are the tenant, and often the unpaid labourer too. Consider what you actually hand over on a typical day.

• You pay rent in attention. Every minute you spend scrolling is inventory a platform sells to advertisers. This is the machinery of the attention economy — your gaze is the crop, and the feed is the field it is harvested from.
• You pay rent in data. Your clicks, locations, searches, and pauses are collected and refined into the raw material that trains ranking systems and, increasingly, artificial intelligence. You produce it; you do not own it.
• You pay rent in free labour. Every review you write, every photo you post, every rating you leave makes the platform more valuable — yet none of that value flows back to you. You improve the lord’s estate for nothing.

The word “free” is doing a lot of quiet work here. These services feel free because no money leaves your bank account. But Varoufakis’s point is that you are paying constantly, in a currency that does not show up on a statement. When a small business must advertise on the same platform where its customers already live, or a maker must sell through the one marketplace buyers actually use, they pay in cash too — a cut on every sale, for the privilege of standing on ground they do not own.

This is why the ownership question behind AI matters so much, and why so many people are now asking who owns AI. The large models being built today are trained on the accumulated output of billions of digital tenants — our writing, our images, our conversations. The commons that produced them is public; the models that capture their value are private. If you have wondered will AI take my job, this is the deeper structural version of the same worry: the tools that may reshape your work are being enclosed by the very few who own the cloud they run on.

Why this is the newest verse of a very old song

Strip away the servers and the metaphor points at something genuinely ancient. To see it clearly, it helps to look at the actual historical event the word “feudalism” gestures toward — not knights and castles, but land and enclosure.

For centuries, English villages farmed common land: fields, pasture, and woodland that no single person owned and everyone could use. Then, across the sixteenth to nineteenth centuries, that common land was steadily fenced off — “enclosed” — and turned into private property. People who had lived off the commons for generations woke up to find the gate locked and a rent demanded. A resource that had been shared was captured, and those who once used it freely became dependent on those who now owned it.

That is an old pattern, visible across ten thousand years: a common resource gets fenced, a few capture the value, and the many who once relied on it are dispossessed. From the granary to the printing press to the cloud, it is the same move, a new machine, every time. Technofeudalism, in that light, is not a shocking break with history. It is the latest verse of a very old song — the digital commons of our attention and data being enclosed exactly the way the fields once were.

A resource that had been shared gets fenced. A few capture the value. The many who relied on it are dispossessed. The same move, a new machine, every time.

You can watch the pattern repeat close to home. When public digital infrastructure like India’s Aadhaar and UPI becomes the ground that daily life runs on, the same questions surface: who owns the rails, who sets the terms, and what happens to the people who have no choice but to use them? The specifics change with each machine — clay tablets, steam engines, fibre-optic cable — but the shape of the story is stubbornly familiar. Someone builds or seizes the chokepoint through which everyone must pass, and from then on the flow of value bends toward whoever holds it.

Seeing technofeudalism as one verse in that longer song does two useful things. It cools the sense of panic — this is not the first time a new technology has been captured, and past captures were eventually challenged and reshaped. And it sharpens the diagnosis, because the recurring pattern tells you exactly where to look: not at the gadget, but at the gate.

The counter-argument: is it really “feudalism,” or just monopoly capitalism?

It would be dishonest to present technofeudalism as settled fact. It is a contested thesis, and many careful economists think the medieval label is more rhetoric than analysis. Their objection deserves a fair hearing, because it is strong.

The core of the counter-argument runs like this. Feudalism, properly defined, was a system of legally fixed hierarchy: serfs were bound to the land, birth determined status, and lords extracted dues through force and law rather than trade. None of that describes Amazon or Google. Their power comes from selling things people voluntarily buy, in markets, for profit, under competition law. Nobody is legally bound to a platform; you can delete the app. What we are seeing, these critics say, is not a new mode of production at all — it is monopoly capitalism, an old and well-understood problem, in which a handful of firms grow large enough to distort markets, extract outsized margins, and squeeze suppliers and workers.

On this view, the “rent” technofeudalism describes is simply what economists have long called monopoly rent: excess profit that flows from market power rather than genuine competition. That is a real harm — but it is a harm capitalism has produced before, from the railroads to Standard Oil, and it was met with antitrust law, not a change of epoch. Reaching for “feudalism,” the critics argue, obscures more than it reveals. It swaps a precise diagnosis — concentrated market power — for a dramatic metaphor, and metaphors make poor policy. Why invent a new system when the old tools of competition enforcement, data rights, and taxation might address the actual problem?

This is a serious position, and you do not have to accept the full technofeudalism thesis to find the underlying worry — a few firms owning the ground everyone else stands on — completely legitimate.

The test that cuts through it

So how should a normal reader decide what to make of all this, without a degree in economics? You do not need to settle whether the word “feudalism” is literally correct. You can ask a sharper, more useful pair of questions — the same test that works on any technology, in any century.

1. What is being enclosed? Find the shared resource that used to be open and is now fenced. With the commons it was land. Today it may be your attention, your data, the reach a business needs to find customers, or the collective human output now being fed into AI. Name the thing that was once accessible and is now owned.
2. Who is dispossessed? Follow the value. Ask who used to benefit from that resource freely, who now pays to access it, and who collects the toll. If a small group owns the gate and everyone else pays to pass, you have found the dispossession — whatever you decide to call the system that produced it.

Run that test and the label matters less than the mechanism. Whether you side with Varoufakis and call it technofeudalism, or with his critics and call it monopoly capitalism, you are pointing at the same underlying event: a valuable common resource being captured, and the many who relied on it being pushed to the margins. The disagreement is about the name. The pattern is not in dispute.

That is the quiet power of learning to see the move for what it is. Once you can spot enclosure — the fence going up around something that used to be shared — you can recognise it in a medieval field, in a monopoly, and in the glowing rectangle in your hand. The machine keeps changing. The move, so far, has not.