Everyone wants their own AI: the new nationalism changing everything

For years, AI has been that fascinating toy used to create absurd images, write texts or automate tasks we used to put off; but while the average user is busy generating surreal videos or asking a chatbot for existential advice, governments have been playing a much more serious game. Because, ultimately, the conversation is no longer just about productivity or creativity, it’s shifting towards something far more uncomfortable: control. And that’s exactly where a term that’s becoming harder to ignore comes in — AI nationalism.

What is AI nationalism?

The idea is fairly simple: each country should develop and control its own artificial intelligence rather than rely on others. In other words, AI is treated as a strategic resource, much like energy or defence. This approach gained traction in 2018, when investor Ian Hogarth introduced the concept of “AI nationalism”, arguing that this technology is not just another tool, but something directly linked to avoiding dependence on other countries. In that sense, the logic is clear: whoever leads in AI doesn’t just gain economically, they also gain political weight and global influence.

And this is no longer theoretical, it’s already happening. This month in the United States, the Pentagon reportedly labelled Anthropic a national security risk after the company refused to remove restrictions on the use of its models for mass surveillance or autonomous weapons. In response, OpenAI signed a $200 million contract with the Department of Defense, effectively placing operational decisions about these systems in government hands. Add to that recent moves to reduce regulation, and the message becomes clear: when AI enters the state arena, it stops being just technology and starts becoming a tool of real control.

There’s also one key factor that explains everything: concentration. Today, the most advanced AI systems are primarily developed in the United States and China, which means the rest of the world depends — to varying degrees — on these two powers. That’s why, as outlets like Financial Times have pointed out, the case of DeepMind — the British lab acquired by Google in 2014 — is often cited as an example of what’s at stake. It wasn’t just a business deal, but also the loss of a key technology that could have given the UK greater independence and influence.

Governments start stepping in

Until recently, all of this sounded like a theoretical debate, but not anymore. In practice, countries like the United States are already making concrete moves around AI. For instance, restrictions on exporting advanced chips to China are not just about trade, they’re part of a broader strategy to slow down technological development in key areas like AI. In other words, technology is increasingly being treated as a matter of security and global competition.

At the same time, countries such as France, India and the United Arab Emirates have announced multi-billion investments to build their own AI infrastructure, aiming to avoid dependence on foreign powers. Meanwhile, tensions are emerging between governments and tech companies when certain limits come into play, particularly around surveillance or military use. All of this points to a clear shift: AI is no longer just about innovation, it’s also about constant negotiation between companies and states — and, at times, pressure.

The upside (in theory)

On paper, AI nationalism makes sense. It allows countries to protect themselves, reduce technological dependence and prevent a handful of large corporations from concentrating too much power in a critical sector.

It could also lead to greater public control over how data is used and how these technologies are developed, something that has long been part of the broader tech debate. Some even see it as a way to accelerate progress, by concentrating investment, talent and infrastructure under a single national strategy. In an ideal scenario — which rarely exists — this could result in more transparency and less opacity.

The downside

The problem appears when theory meets reality. Put simply, if it’s already unsettling to see big tech accumulating so much power, it’s even more concerning to imagine governments with full access to advanced AI systems, especially when used for surveillance, control or military purposes — issues that have been widely debated.

There’s also another obvious friction: speed. AI evolves rapidly and globally, while governments tend to move much more slowly, meaning excessive or poorly timed regulation can end up slowing down the very progress it aims to support. And then there’s the cost, because developing or controlling this technology is extremely expensive, to the point of potentially impacting entire national economies.

Dystopia? Not quite

It’s easy to slip into apocalyptic thinking, but it’s not that simple. We’re not living in a real-time sci-fi film or heading inevitably towards total control, although there is a clear shift: AI is no longer a neutral tool and is becoming embedded in traditional power dynamics. In the end, this isn’t entirely new — every major technological breakthrough tends to become a strategic resource. The difference is that this one learns, evolves constantly and relies on global, real-time data, making it much harder to manage.

So the real question is no longer whether AI is good or bad, but who controls it, for what purpose and under what limits. Too much power in private hands can lead to monopolies that are hard to regulate; too much in public hands can open the door to even greater risks, such as abuse or lack of accountability. For now, there’s no clear balance, and it will likely take time to find one. In the meantime, AI keeps advancing — and with it, a slightly uncomfortable feeling that the future isn’t just being built, it’s being contested

Spotify wants users to create AI-powered remixes on its platform.