Throughout human history, power had a footprint. A nation has borders. An institution has walls. A ruler has a face. When something threatened you—be it a king, an army, or a corporation—you could point to it. You could see where it lived.

Networks operate differently. Their power doesn’t come from solid structures. It comes from connection. And unlike nations or people… networks can easily stay invisible.

Now, occasionally, a network wants to be seen. Consider the digital influencer—the highly visible hub at the center of millions of followers. Here, the network wears a human face. It feels personal. It feels transparent.

But that visible edge is an illusion. The real danger lies in what stays below the surface.

We see this pattern playing out again and again across recent history:

• Look at the anti-Trump movement that erupted a decade ago. Donald Trump was a household name and a public figure for decades, yet almost overnight, a deeply coordinated network mobilized to challenge him.
• Look at COVID-19, which saw an unprecedented global response instantly organized across borders, institutions, and platforms.
• Look at the war in Ukraine, where an organic-looking wave of support swept through the U.S., cross-cutting political lines even when most citizens barely understood the regional dynamics.
• Or consider the pro-Palestinian movement after October 7th, which appeared to burst forth organically, yet relied on sophisticated, pre-existing organizational structures and digital architecture ready to activate at a moment’s notice.

Up until now, these invisible webs were driven by human decisions—people organizing other people behind closed doors. But that paradigm is shifting radically.

Enter the age of artificial intelligence. We are no longer just dealing with networks of humans connecting with humans. We are living alongside synthetic networks.

Today, human beings build these AI networks. Tomorrow, these networks may create themselves—and run on autonomous “agents” directing fleets of “sub-agents” to accomplish tasks, all without any human intervention required in the chain.

These AI-driven networks aren’t just built to promote ideologies or manufacture political narratives. In fact, many are built for far more mundane, practical reasons: simply helping businesses become hyper-efficient by empowering agents to handle complex workflows.

An agentic AI network doesn’t need a malicious ideology to spin out of control. Sometimes it’s just human greed or laziness, poor system design, or flimsy guardrails in pursuit of corporate efficiency. Other times, the technology simply gets so advanced that it develops its own implicit moral code—one that subverts the human instructions it was given because it decides its route is superior.

Whatever the cause—whether driven by geopolitical agendas, corporate optimization, or machine-derived ethics—the effect is the same: the carefully crafted nations, institutions, and laws that have governed human behavior for centuries can be completely disabled the moment we hand real control over to the machine.

AI allows for the instant creation of ghost architectures—thousands of synthetic personas, automated accounts, and algorithmic loops, all acting in perfect synchronization. They don’t just react to reality; they manufacture it. They buy assets, sway elections, destroy reputations, and engineer outrage out of thin air.

Instead of being passively acted upon by these hidden systems, we need to step up and take back control.

My new book, Real-World Geopolitics, available now on Amazon, is an effort to wake people up to the extraordinary power of these networks and to support a robust conversation about them in the public space.