Mutually Assured Destruction (MAD)
The nuclear standoff in which each side can destroy the other even after absorbing a surprise attack, so striking first buys nothing.
Mutually assured destruction is the condition in which both sides of a nuclear standoff can absorb a full surprise attack and still annihilate the attacker afterward. Once that is true, hitting first buys you nothing. You do not disarm your enemy, you only guarantee your own destruction. Almost every nuclear decision of the past sixty years has been about making that condition permanent, and about making sure the other side knows it is permanent.
Bernard Brodie saw it in 1946, a year after Hiroshima, when he wrote that the purpose of a military establishment had until then been to win wars and from now on would be to prevent them. The technology to make that real took another twenty years.
Making it work
The requirement is a second strike: forces that survive being attacked. That single demand drove most of the Cold War arms race. Missiles went into hardened concrete silos. Bombers sat on runway alert. Above all, warheads went to sea in ballistic missile submarines that nobody can reliably find, so that even a flawless first strike leaves hundreds of warheads loose under the ocean. The American triad of land missiles, bombers, and submarines exists so that no single breakthrough in detection or accuracy can take out the whole force at once.
Robert McNamara turned it into a number. In the mid-1960s the Pentagon defined assured destruction as the ability, after absorbing a Soviet first strike, to destroy roughly a quarter of the Soviet population and half of its industry. Beyond that, extra weapons add nothing, which became the argument for capping arsenals. The acronym MAD was coined by a critic, Donald Brennan of the Hudson Institute, who thought deliberately keeping your own population undefended was insane. The name stuck because both camps found it useful.
The strangest consequence is that missile defense is destabilizing. A country shielded from retaliation is a country that might risk striking first, so a shield makes the other side build more missiles to overwhelm it. That is why the 1972 ABM Treaty limited defenses instead of offenses. Both superpowers agreed to stay vulnerable on purpose. The United States withdrew from it in 2002.
What it does not cover
The logic has holes, and all of them are still open.
It assumes leaders act rationally and that warning systems work. In 1983 a Soviet early warning satellite reported American missiles inbound. The duty officer, Stanislav Petrov, judged it a malfunction and did not pass it up the chain. He was right. Deterrence survived on one man's guess.
It says nothing useful about small nuclear use. A single tactical warhead on a battlefield does not trigger mutual annihilation, and no side has a clear doctrine for what happens next. That gap is precisely where Russian nuclear signaling over Ukraine has been aimed.
And it was designed for two players. With nine nuclear-armed states it is far harder. India deters Pakistan and China simultaneously, China deters the United States and India, and anything any of them builds unbalances two relationships at once.
Where it stands
The machinery that managed all this is coming apart. The INF Treaty ended in 2019. New START, the last agreement limiting American and Russian deployed strategic weapons, expired on 5 February 2026 with nothing to replace it, leaving the two largest arsenals without a binding limit for the first time in more than half a century. Both governments say they will respect the old ceilings for now, which is a statement rather than a treaty. China meanwhile is expanding fast enough to become a third full-sized player, and the United States is building missile defenses again.
None of that repeals the arithmetic. As long as a few hundred warheads survive any attack, going first is still suicide. What is disappearing is the inspections, notifications, and limits that let each side verify what the other actually has, and deterrence runs on belief.