AI & States: Government, Institutions, and Public Power

When artificial intelligence enters the public sphere, it does not arrive as a neutral upgrade. It arrives as a pressure test. The State is the only actor that must be both guardian and user, both regulator and practitioner, both the one who sets limits and the one who needs capability. That tension is not a footnote. It is the story.

This category is devoted to the relationship between AI and public power: governments, administrations, courts, regulators, central banks, supranational authorities, and the institutional ecosystems that hold modern societies together. The question is not whether the State will “adopt AI”. The State is already being reshaped by it. The question is what kind of State will emerge when intelligence becomes infrastructure, when decision-making can be accelerated, outsourced, or embedded in systems that few fully understand.

In the articles published here, AI is treated as a political and institutional variable before it is treated as a technical one. Because the most consequential effects do not happen in laboratories, they happen where authority is exercised: in procurement choices, in regulatory definitions, in the design of public services, in security doctrine, in the management of borders, in the automation of compliance, in the quiet migration of discretion from humans to systems.

This is where the debate becomes serious for decision-makers and entrepreneurs alike. Firms operate within institutional environments. Markets rely on legal certainty. Capital depends on predictable rules. When AI starts to alter how rules are written, enforced, and operationalized, every strategic landscape changes. Sometimes subtly, sometimes abruptly.

A central theme of this category is sovereignty, not as a slogan, but as an operational problem. AI systems depend on compute, data, and supply chains that are unevenly distributed across the world. They depend on infrastructure that is often owned, hosted, or governed outside national borders. The result is a new kind of dependency that does not look like old industrial dependence. It looks like cognitive dependence. States can legislate, but can they execute. They can declare principles, but can they implement. They can set standards, but can they shape the stack that makes standards real.

Another theme is legitimacy. Institutions derive legitimacy from transparency, accountability, and due process. AI introduces opacity, scale, and speed. It can improve efficiency, but efficiency is not a constitutional principle. The State cannot be evaluated like a corporation. It is judged by different metrics: fairness, rights, proportionality, contestability. When algorithmic systems start to mediate access to services, benefits, permits, justice, or security, the institutional question is no longer technological. It becomes democratic.

This category is also where Europe’s position becomes visible. The European instinct is to regulate, to define boundaries, to protect citizens and markets from uncontrolled concentration of power. The American instinct has often been to scale first and govern later. The Chinese model ties AI to state capacity and strategic coordination. These are not cultural differences. They are institutional architectures. And entrepreneurs operating across borders increasingly need to understand them as part of the operating environment, not as distant political theatre.

QUI MILANO hosts this work as a deliberate editorial stance. Not to narrate a territory, but to read the world from one of its decisive European nodes. Milan is a place where corporate strategy, financial capital, regulatory signals, and public policy debates intersect. From this vantage point, the Observatory treats the State not as a national anecdote, but as a global actor under structural transformation. It looks at how public power is being retooled, constrained, or expanded by AI, and what that implies for markets, governance, and institutional stability.

The tone of this category remains intentionally sober. It does not dramatize. It does not promise. It does not indulge in the fashionable idea that technology will “solve” politics. It assumes the opposite: that technology intensifies politics by making choices unavoidable. AI does not remove discretion. It relocates it. It does not eliminate power. It reorganizes it.

And in a world where the rules of the game are being rewritten by code, every serious reader eventually asks the same question, quietly, before any others: who is writing the rules, and on whose infrastructure.