Why countries no longer feel secure under a single nuclear umbrella

By Rebecca Lissner, a senior fellow for U.S. foreign policy at the Council on Foreign Relations, and Erin D. Dumbacher, the Stanton nuclear security senior fellow at the Council on Foreign Relations.

In 1960, a young U.S. senator named John F. Kennedy spoke almost casually about the future of nuclear weapons. Within a few years, he warned, the world might see as many as 20 nuclear-armed states. At the time, this did not sound alarmist—it reflected a widely shared assumption about where history was heading.

More than six decades later, only nine countries possess nuclear weapons. For a long time, it seemed that the worst-case scenario had been contained. Today, that confidence is beginning to fade.

What is happening

Across policy and security circles, there is growing discussion about the transformation of the global nuclear order. Arms control agreements are eroding, the nonproliferation regime is under strain, and U.S. allies are increasingly uncertain about the long-term reliability of American security guarantees.

These dynamics have been examined in recent analytical work in Foreign Policy by Rebecca Lissner and Erin D. Dumbacher. But beyond the specifics, the broader shift deserves closer attention.

What really matters here

If we strip away the details, the core change is not about rising aggression or renewed appetite for nuclear power. It is about the loss of confidence.

For decades, the global nuclear system rested on a simple psychological foundation: when things truly mattered, there was a central guarantor willing and able to bear the ultimate risk. That belief reduced incentives for proliferation and anchored deterrence.

As that confidence weakens, everything else begins to move.

Why states are acting this way

In situations of uncertainty, actors rarely seek the best possible outcome. They focus instead on avoiding the worst one.

For non-nuclear states, the worst-case scenario is not strategic disadvantage—it is standing alone in a moment of existential threat, relying on assurances that may not hold. In that context, developing an independent nuclear arsenal is often too risky, costly, and destabilizing. A more attractive option is to look for additional guarantees, alternative backstops, or parallel security arrangements.

This is not rebellion against the old order.
It is a form of insurance.

How this changes the logic of nuclear security

The emerging trend does not necessarily point toward rapid nuclear proliferation. A more likely outcome is the distribution of nuclear responsibility.

Security guarantees multiply. Decision-making becomes more dispersed. The system gains redundancy—but loses clarity.

Such an arrangement may appear stable in calm periods. In moments of crisis, however, it becomes harder to read, slower to manage, and more vulnerable to miscalculation.

What this means in the long run

The world is entering a phase in which security depends less on trust and more on fallback options. This reduces the risk of sudden collapse but increases the risk of gradual erosion and error.

Nuclear weapons remain tools of deterrence. What is changing is the architecture that governs how deterrence works—and who is expected to carry its weight.

Previous
Previous

Energy Outlook 2026: when abundance meets infrastructure scarcity

Next
Next

How China wins the future: what this means for strategy, risk, and competitive position