Oligarchy

Oligarchy literally means “rule of the few”. The word comes from Greek: “oligoi” means few, “archein” means to rule. It refers to a state in which a small group makes the important decisions. This group is not elected, and the majority can only get rid of it with great difficulty. Its power usually rests on money, property, origin, or on control over something everyone else needs. The counter-term is democracy, in which power emanates from the totality of citizens and is regularly reassigned.

Why a few companies decide over the AI world

In the tech industry, the word has appeared more and more frequently in recent years. The reason is an imbalance: building modern AI systems is extremely expensive. It requires special chips, huge data centers, and enormous amounts of electricity. The costs for a single large model easily run into the hundreds of millions. A university or a startup cannot raise that.

What remains is a handful of corporations, mainly from the USA and China. They decide which systems are developed and who is allowed to use them. They set the prices and determine the rules by which their systems respond. Critics call this an oligarchy of computing power. The accusation is: a technology that affects millions of people is controlled by a very small circle.

An important distinction is needed here. A monopoly means that a single provider dominates a market. An oligopoly means that a few providers share the market among themselves. Oligarchy is the broader, more political term. It describes not just market shares, but influence over laws, public debates, and entire societies.

Where the power of the small group comes from

Oligarchic power rarely arises from a single decision. It grows because advantages reinforce themselves. Whoever has a lot of money can hire the best experts. Whoever has the best experts builds better products. Better products bring in more money again. Experts call this cycle a feedback effect.

Added to this is control over bottlenecks. A bottleneck is something everyone needs and only a few can supply. In AI, this mainly means specialized chips for training models. A single manufacturer covers the largest part of the world market here. Whoever controls this access also indirectly decides who is even allowed to play.

Another building block is political influence. Large corporations maintain offices in Brussels and Washington and advise authorities on new laws. This is legal and often sensible, because genuine expertise sits there. It becomes problematic when rules end up protecting those who helped write them. New competitors then fail due to requirements that are easy for the big players to meet.

The term in headlines and debates

In business news, you encounter the word in two contexts. The first is politics, for example in reports about Russia or other countries where a few very wealthy individuals co-govern the state. The second is technology. There, commentators write of an “AI oligarchy” or of “tech oligarchs” when criticizing the power of individual company leaders.

In very concrete terms, you notice more of this than it seems. Which videos a platform suggests to you, which search results appear at the top, which chatbot is built into your phone: these are decided by the rules within the systems of a few providers. If one of these companies changes its settings, something changes for millions of people at once. This is precisely the reach critics mean when they speak of concentration of power.

A common misconception is that “oligarchy” is simply a slur for large companies. That misses the point. The term describes a structure, namely lasting power without genuine external control. Countermeasures do exist: antitrust proceedings, openly accessible models, and independent audits. Whether they work is one of the open questions of the coming years.