Recently, Palantir chief executive Alex Karp predicted that artificial intelligence companies could be nationalised within the next few years.

Perhaps he is right. Perhaps he is wrong.

But I suspect he is asking the wrong question.

The real question is not whether governments will own artificial intelligence, it is whether ownership will even matter.

After all, what is the practical difference between a government owning the most powerful information system in human history and a government exerting influence over it?

That may sound like a radical proposition. In reality, it is merely the latest chapter in a very old story.

For most of human history, power meant controlling territory. Kings fought wars over land because land generated wealth. Later, power came to mean controlling industry. Nations competed for factories, resources and productive capacity because economic strength determined geopolitical strength.

Today, power increasingly means controlling information.

The institutions that shape what people know have always exercised extraordinary influence. Churches controlled scripture. Newspapers controlled the news. Universities shaped accepted knowledge. Television networks determined which stories entered the national conversation and which did not.

Artificial intelligence represents something fundamentally different.

It is not simply another media platform or search engine. Increasingly, it is becoming a teacher, researcher, adviser, librarian, and editor all at once. Millions of people already use AI to answer questions, explain complex issues, summarise events and conduct research. For many, it is becoming the primary interface through which they understand the world.

That should concern us all.

Not because AI is inherently dangerous, but because concentrated authority always is.

The debate surrounding social media over the past decade offers a glimpse of what lies ahead. Governments have become deeply involved in discussions about misinformation, extremism, hate speech and harmful content. Technology companies have responded with moderation systems and increasingly complex rules governing what can and cannot be said.

Whether one agrees with those policies is almost beside the point.

The important observation is that the relationship already exists. Governments influence platforms. Platforms influence public discourse. Public discourse influences political outcomes.

Artificial intelligence takes that dynamic to an entirely new level.

A social media platform can influence what people see. An AI system may influence how they understand it.

That distinction matters.

The nightmare scenario is not an AI that lies. Lies are usually discovered eventually.

The nightmare scenario is an AI that quietly narrows reality.

One fact omitted here. One perspective excluded there. One argument treated as respectable. Another treated as dangerous. Not enough to be noticed. Just enough to shape perception.

Previous generations feared governments controlling what citizens could say. The next generation may confront institutions capable of determining what citizens know – or perhaps more accurately, what they never get the opportunity to know.

For centuries, free societies operated on a simple assumption: truth emerged through competition. Competing newspapers, competing political parties, competing institutions, and competing ideas. The process was messy and imperfect, but it prevented any single authority from becoming the sole arbiter of reality.

That is why the nationalisation debate misses the point.

The danger posed by advanced AI is not necessarily government control. Nor is it necessarily corporate control.

The danger is concentrated control.

If a handful of organisations become the primary source through which billions of people access knowledge, ownership becomes a secondary concern. The more important question is who establishes the rules.

Who decides what constitutes misinformation? Who determines what is hateful? Who defines extremism? Who decides which perspectives remain legitimate?

Every society must answer those questions. The concern is that fewer and fewer people may be involved in answering them.

Perhaps that is the question governments should be asking now.

Not who owns artificial intelligence. Not how much tax it should pay. Not even how many jobs it may replace.

But how free societies preserve freedom itself when information is increasingly mediated by machines.

How do we ensure citizens retain access to competing viewpoints?

How do we prevent governments, corporations or activist groups from quietly narrowing the boundaries of acceptable thought?

How do we preserve freedom of inquiry in an age of artificial intelligence?

Competition is certainly part of the answer.

But it is not the whole answer.

The larger challenge is ensuring that no institution acquires the power to decide what everyone else is allowed to know.

These questions are no longer theoretical.

The rules governing artificial intelligence are being written right now.

The technology is advancing at extraordinary speed.

The public debate is not.

And that should concern us all.

Because if governments are not asking these questions now, they may discover the answers only after somebody else has already decided them.