The use of nuclear weapons presents a growing risk in the world today, arising from the increasing number of nuclear-armed countries—now nine and counting. The risk also comes from a less appreciated source: the development of new war technologies. AI, cyber, hypersonic missiles, autonomy, space weapons, network hackers, drones, lasers, robot ships, and submarines all interweave with this thickening nuclear context in complex, poorly understood ways.
Nuclear weapons were used in the first nuclear age as the major threat of the Cold War, and they are being used now, in a second nuclear age. They are being used as threats in the same way a bank robber uses his gun. One need not fire a nuclear weapon to use it.
The relationship between nuclear weapons and the new war machines is often overlooked.
This is because the topic is difficult, unwelcome, and sobering. It is easier to focus on an AI-driven military revolution signaled by advances in computer chips, autonomy, and hypersonic missiles.
Here, nuclear weapons are ignored. Or they are filed to an annex labeled “to be opened only in certain largely unimaginable and highly unlikely circumstances.”
Any realistic assessment of the new technologies requires a nuclear frame of reference because it will powerfully influence how—and even if—these new technologies are used. Their impressive performance will be shaped as much by this nuclear context as by their technological potential.
The New Strategic Environment
There are two technology enterprises in the global security environment: the new war machines, and “old” nuclear weapons. New war machines are cutting edge. They bring career advancement, funding, and excitement in the West. Nuclear weapons do not.
The U.S. military, most of it, treats the bomb as a necessary evil. Most officers choose to build their careers around the new war machines. The number of currently serving military officers who have ever seen or handled a nuclear warhead is small. The number who have flown off a carrier with a nuclear weapon is zero.
Something more is that intellectual capital concerning nuclear issues has been hollowed out. The topic no longer draws the creatives, the brains, that it once did. Until recently, many political leaders encouraged this brain drain, in order to stigmatize the bomb, to set an example for other countries that it didn’t really matter very much. The United States has nuclear weapons, yes, but only because others have them. So, the argument went, other countries should not be interested in getting “the bomb” for themselves.
Unfortunately, nuclear weapons are not looked at in this way by other countries. To China, Russia, India, Pakistan, North Korea, Israel, and Iran, nuclear weapons are not outmoded. My guess is that they are against nuclear war, too, like a bank robber who doesn’t want a shootout.
Yet they are putting an enormous effort into their nuclear programs. They are thinking up ways to incorporate nuclear weapons into their broader national strategies, as the United States once did but no longer does. Their intellectual capital grows while that of the United States has stagnated.
For example, China’s military does not see it this way. China’s People’s Liberation Army (PLA) is building the new war machines and nuclear weapons.
Something a lot more subtle and nuanced is going on here. In fact, when the Commander of U.S. Strategic Command, Admiral Charles Richard, described the Chinese buildup, he hinted at this. He called it a strategic, not a nuclear breakout. This astute description gets to the unwelcome subject of the purpose of these weapons. People, including many experts, say that the only purpose of nuclear weapons is to deter other nuclear weapons.
But a strategic breakout says something different: that there are uses beyond narrow deterrence. Purposes beyond “deterrence of other nuclear weapons” often becomes an emotional topic. Even so, nuclear weapons must be examined in a clinical way, if for no other reason than that our main rivals in China, Russia, and North Korea have them, and Iran aspires to them. Moreover, there really are other uses of the bomb beyond narrow deterrence.
One use is to drive a wedge between the United States and its allies over the danger of nuclear war. Another is for brinkmanship, to ratchet up risk in a crisis. Still another is to threaten high-tech conventional war in a thick nuclear context to dissuade the United States from fighting at all.
One more use of nuclear weapons is to bolster the sense that a non-White, post-colonial order is taking shape to stand up to Western dominance and the West’s self-serving laws and grandiose declarations of principle. This new order breaks the Western monopoly grip on the bomb and other military technologies. It suggests too that it is the United States that is on the wrong side of history.
All of these uses of nuclear weapons are on the table now. All go beyond the narrow framing of stability defined in terms of second-strike surviving forces, which has so dominated thinking in the United States on nuclear strategy.
Counterforce in the Second Nuclear Age
Overly narrow framing of the use of nuclear weapons was a problem in the past. But the new war machines now give an ability to destroy others’ nuclear forces with conventional weapons.
Call it conventional counterforce. It combines information from AI, drones, target recognition, satellites, undersea sensors, and cyber hacks with the new war machines, hypersonic missiles, stealth, autonomy, and others.
The result is a conventional kill network that can destroy nearly any target. Moreover, it operates even against moving targets: mobile nuclear missiles, ships, satellites, command posts. This new capability has far-reaching impacts. Yet many studies and war games treat it too narrowly.
I have found that studies and war games frequently have an air of unreality about them. They begin with a strike by one side and a counterstrike by the other. They usually go on to calculate the surviving forces on each side after one or two additional back and forth salvos. Then the game ends. The problem, the unreality, comes from not including how nuclear threats and counter threats play into political decisions.
It’s like analyzing a bank robbery shootout by the caliber, accuracy, and number of bullets fired by the robbers and the guards. These matter, certainly, but real bank robberies are not like this. They go wrong. They get messy, with panicked decisions that make little sense. Things happen that the robber never considered beforehand, and the bank guards, too. There’s no icy rational calculation of damage. Real wars go on much longer than two or three salvos. Years, in fact, when the studies and games predicted they would be over quickly once each side cooly calculated the damages and the gains. Vietnam, Iraq, Afghanistan, and Ukraine are all examples. Will Taiwan be a long war as well?
Take North Korea. The United States threatens to destroy North Korea if they use nuclear weapons. In Cold War terminology, this is a countervalue threat. This is fine, maybe, as a declaratory policy. It shows that the United States is all about deterrence.
Yet it leaves out threats and counterthreats that are likely to arise. The highest U.S. priority must be the hunt for North Korea’s nuclear weapons, the bulk of which are on mobile land-based missiles. Destruction of cities and infrastructure directs attention and resources away from this.
If war between the United States and North Korea breaks out, the United States is likely, initially at least, to avoid strikes on civilians, both to not provoke North Korean nuclear retaliation and to destroy its nuclear weapons. This is also to limit damage to allies and to end a war that, in all probability, neither side had wanted or intended. No U.S. leader wants to go down in history as the second president to kill hundreds of thousands, or now millions, of Asians with atomic bombs.
In an intense crisis, North Korea is likely to disperse its mobile nuclear launchers, unlock its warheads, and perhaps pre-delegate launch authority—creating a dangerous situation. What drives the danger is less a US nuclear threat than the enhanced conventional counterforce threat. So conventional counterforce against a nuclear weapon state brings important decision points to the fore. The United States would be unwise to overlook these—far better to understand them and to play them out in studies and games.
But in-depth study and wargaming is needed for another reason, too: to answer the question a U.S. president will ask: “Am I actually authorizing a conventional start to a nuclear war?” This is not an unreasonable question. There are foolish answers to it, ones that I have seen in studies and war games. “North Korea will not go nuclear because they know what would happen if they did.” Or, “Sir, our AI models say we can destroy 60 percent of their missiles.”
These are not discussions about a botched exit from Saigon or Kabul here. These are presidential decision points that would result in the deaths of millions of people with vast consequences. These decision points, if mishandled, could lead to either accommodation or paralysis.
Either would be a national security disaster even if nuclear war is avoided.
Strategist Herman Kahn defined a realistic war plan as one where the president didn’t throw you out of the office. If we fail this elementary test of realism, all the AI, cyber, and hypersonic missiles in the world aren’t going to matter. Declaratory peacetime statements and doctrine will be tossed out the window in a real crisis.
The first question any president will ask is where a conventional counterforce strike on a nuclear state goes next. Failing to point out that the purpose of the said attack is to disarm the enemy of his nuclear weapons is irresponsible and untenable.
Better options are essential, otherwise the alternative is accommodation or a nuclear war that the United States is ill equipped to fight.
 Aaron Mehta, “STRATCOM Chief Warns of Chinese Strategic Breakout,” Breaking Defense, Aug. 12, 2021.
 There are exceptions to this. A recent Center for Strategic and International Studies (CSIS) war game of a Chinese attack on Taiwan allowed for prohibition of US conventional attacks on the Chinese mainland because of fears of nuclear escalation. But it did not play political decisions or nuclear threats of the kind emphasized here. See Mark F. Cancian, Matthew Cancian, and Eric Hegenbotham, “The First Battle of the Next War, Wargaming a Chinese Invasion of Taiwan,” Report of the CSIS International Security Program, Jan. 2023
 For more on this see Paul Bracken, The Hunt for Mobile Missiles, Nuclear Weapons, AI, and the New Arms Race (Philadelphia, PA: Foreign Policy Research Institute, 2020).
The article is excerpted from the author’s broader article entitled “Navies in the Second Nuclear Age” published by the Foreign Policy Research Institute and is posted with the author’s permission.