Arms control used to set the outer limits of what the nuclear powers built. Since New START expired in February nothing does, and decisions taken separately in Washington are starting to add up. None of them on its own is new in kind. What is new is that they are happening at once, with no treaty framework to cap the response.
More options for the president
The Pentagon is reviewing the guidance that tells the military how nuclear weapons could be used. Elbridge Colby, the under secretary of defense for policy, described it in August as a "relook at the employment guidance" with US Strategic Command, and said it would not be published. It is being run as a strategy review, not a formal Nuclear Posture Review, which Colby had told the Senate in March would not be conducted.1
Officials say the aim is "credible, rational nuclear options" for regional wars, so that China or Russia do not conclude the US would hesitate to respond to a battlefield nuclear strike over Taiwan or Ukraine. Reporting says the review gives more weight to limited and lower-yield options; no new systems have been confirmed.1 Critics see it differently. Joe Cirincione, writing in the Bulletin of the Atomic Scientists, calls it a plan that "will give President Donald Trump more ways to start a nuclear war", and argues that war games involving tactical nuclear use have almost always escalated.2
The pits
Plutonium pits are the cores of nuclear warheads. The US has not made them at scale since 1989. The National Nuclear Security Administration's goal is 80 a year, split between Los Alamos in New Mexico and a converted facility at the Savannah River Site in South Carolina. On 15 October the Energy Department approved the cost and schedule baseline for the main Savannah River building, clearing full construction.3 A draft environmental impact statement published in April assesses the programme.3
The programme is contested on two grounds. On cost and schedule, critics say the Savannah River conversion is years late, and Congress has questioned overruns at budget hearings.4 On need, two scientists writing in the Bulletin this week argue that the agency is "cutting safety corners and risking nuclear instability" to fix a problem, ageing pits, that independent studies have not found.5 The agency's position is that new pits are needed to sustain and modernise the arsenal.3
Golden Dome
The administration's Golden Dome missile-defence plan was announced at $175 billion. In May the Congressional Budget Office estimated that a notional version would cost about $1.2 trillion over 20 years; without the space layer, about $448 billion. About 60% of that is a constellation of roughly 7,800 space-based interceptors, which would need about 1,600 replacements a year. Even then, the CBO found, it could engage only about ten missiles at once and "would not … be able to fully counter a large attack of the sort that Russia or China might be able to launch."6 Physicists Lisbeth Gronlund and David Wright of MIT reach a similar conclusion on technical grounds.7
General Michael Guetlein, who runs the programme, disputes the estimate and puts the cost at $185 billion; the CBO says the gap reflects differences in scope, time frame and assumptions.6
Missile defence matters for arms racing because it gives adversaries a reason to build more. China has said as much: its own state media now links its new missile silos to the need for a second strike that can get through US defences (see China's arsenal steps into the open).
Weapons in orbit
On 14 September the Secretary of the Air Force, Troy Meink, said publicly for the first time that the US has "on-orbit space control weapons". He gave no details of their type or number. The next day the chief of space operations described them as able to defend against space-enabled attacks, which analysts read as including other countries' military satellites.8 Satellites provide the early warning and communications on which nuclear command depends, so weapons that can reach them bear directly on nuclear stability.
The case for, and the spiral
The administration's argument is that the US now faces two nuclear peers for the first time, that China is expanding fast, and that deterrence needs more options and a modern production base. That argument is not frivolous: China's stockpile has grown from about 350 to 620 warheads since 2020.9 The counter-argument is the classic security dilemma. Each step that one side takes for its own security is read by the other as a threat, and without treaties there is no mechanism to stop the steps feeding each other.