Trust, AI, and quantum technologies in the new logic of state power

A state may have secure borders, an independent government and a strong economy, yet still depend on systems it neither owns nor controls. Its transactions may run through cryptographic infrastructure designed elsewhere; its critical services may rely on foreign cloud platforms; and its industrial base may depend on semiconductors produced in a handful of locations beyond its reach. Formal sovereignty and functional autonomy have diverged, and that gap is now a central strategic challenge for governments.

The gap is measurable. A 2025 review of 775 data center projects outside the United States found that American firms operate close to half of them when weighted by investment value, used by the researchers as a proxy for compute capacity. Combined with the already substantial concentration of compute physically located in the United States, the study illustrates how jurisdiction can extend well beyond where infrastructure is built. A separate study of leading public cloud providers reaches the same conclusion from another direction: access to advanced public cloud AI compute remains concentrated in a relatively small number of countries, and fewer still host infrastructure capable of supporting advanced AI training.

Power has always followed control over critical infrastructure. What has changed is the nature of that infrastructure: increasingly computational, highly concentrated, and dependent on supply chains and operators that cross national jurisdictions.

Sovereignty, in other words, is becoming computational: the ability of a state to maintain meaningful control over the technologies on which its strategic decision-making depends. This is not a call for autarky, which is unavailable to most countries anyway. It is a question of governance rather than isolation, and of recognizing that a nation can lose strategic autonomy without losing an inch of territory.

Two developments make this urgent. The first is speed. Machine-learning systems can identify vulnerabilities faster than human analysts; cyber operations unfold at machine speed, and automated defenses increasingly contend with automated attacks. As decision windows shrink, institutions may face greater pressure to act before information is complete. The risk is not simply that machines become uncontrollable, but that human institutions must govern environments operating at speeds for which their decision processes were never designed.

The second development concerns what states are ultimately trying to protect: trust. Trust has always underpinned political order. Governments function because citizens trust institutions, economies because participants trust contracts, alliances because states retain sufficient confidence in treaties and deterrence. That trust was once produced primarily by institutions: governments issued currency, courts enforced contracts; diplomatic frameworks reduced uncertainty between states.

Increasingly, trust is also produced by computation. Every secure transaction, identity verification, encrypted message and software update depends on cryptographic mechanisms that establish trust between parties who may never meet or share a jurisdiction. Trust has become infrastructure, as foundational to modern societies as ports, roads and power grids were to industrial ones. Cryptography, once treated as a narrow technical discipline, is better understood as the architecture of confidence in digital life.

Quantum technologies raise these stakes considerably. Today's quantum computers cannot break the widely deployed public-key cryptography protecting financial systems, government communications and critical infrastructure, and there is considerable uncertainty over when a cryptographically relevant quantum computer will emerge. But the risk does not wait for the machine. Encrypted data can be harvested now and held until decryption becomes feasible, which places information transmitted today, and meant to stay protected for decades, at risk now.

Governments have begun to treat this as national policy rather than technical housekeeping. In June 2026, the United Stats issued an executive order naming that collection risk explicitly and setting migration deadlines for federal High Value Assets and high-impact systems: post-quantum key establishment by the end of 2030 and digital signatures by the end of 2031, alongside measures intended to drive post-quantum requirements into federal procurement. France's national cybersecurity agency is moving to introduce post-quantum requirements for products entering its qualification process from 2027. The UAE's national post-quantum migration program similarly begins with visibility: identifying vulnerable cryptographic assets, establishing baselines and prioritizing migration pathways across critical systems.

Post-quantum migration is therefore not simply a technical upgrade. It is a sovereignty transition: an effort to preserve trusted decision-making in an environment where the mathematical foundations of digital trust may no longer hold.

Three priorities follow for governments.

First, treat artificial intelligence, semiconductors and quantum technologies as strategic capabilities rather than simply economic sectors, with the planning horizons and investment patience required.

Second, elevate national post-quantum migration to a matter of sovereign resilience. Start with cryptographic discovery and set migration timelines against the lifespan of the data being protected rather than the expected arrival date of the threat.

Third, make sovereignty a procurement standard. The European Commission's Cloud Sovereignty Framework grades suppliers across eight dimensions, including legal and jurisdictional control, data and AI, operational control, supply chains, technology, security and compliance. Whatever one makes of the particulars, it turns an abstraction into something a contract can test.

Dependencies are rarely national in scope, and remedies will not be either. Trusted supply chains and shared standards matter more than any single national buy-out.

Computation does not erase geography, military power or economic weight, but it does introduce a new layer of strategic dependency. States that secure the computational foundations of trust will preserve their autonomy. Those that do not may discover that the future of state power is decided not only at borders, but within the computational systems upon which modern life depends.