Too Close for Comfort: Cases of Near Nuclear Use and Options for Policy

Sasan Aghlani, Chatham House, outlines just how close the world has come to the inadvertent use of nuclear weapons in the last 60 years – and suggests some immediate measures to reduce the risks.

Article from SGR Newsletter no.43; online publication: 3 April 2015

Download pdf of article [52KB]

Nuclear weapons have not been detonated in conflict since 1945. There is a danger however of becoming too complacent about this record of non-nuclear use. If risk is defined as probability × consequences, the risk of nuclear use is much higher than we have long assumed.

A recent Chatham House report documents 13 instances between 1962 and 2002 where nuclear weapons were almost inadvertently used due to miscalculation, miscommunication, or technical errors.[1] What prevented their use on many of these occasions was the ‘human judgement factor’ – intervention of individuals who, based on prudent assessment of situations and against protocol, either refused to authorise a nuclear strike or relay information that would likely have led to the use of nuclear weapons.

Decision Making under Pressure

A recurring theme in the report is that those involved in ‘command and control’ are under great psychological pressure when making decisions regarding nuclear use, chiefly due to the short window for action. In one such case a research rocket was mistaken for a Trident missile in 1995 and Russian President Boris Yeltsin had only minutes to decide whether to launch a retaliatory strike against the United States. He delayed his decision for as long as possible while following the rocket’s trajectory, talking over the phone with the possessor of the second ‘nuclear briefcase’ until it became clear that the rocket would land outside of Russian territory.[2]

In another example, in 1979, US National Security Advisor Zbigniew Brzezinski received a call from a General at US Strategic Command stating that the Soviet Union had launched 220 nuclear missiles at the United States. A minute before informing the President that the United States was under attack, he received another phone call stating that the alert was caused by a faulty computer chip.[3] It was Brzezinski’s decision to delay his call to the President which proved decisive.

Political Climate

Many of the cases examined in the Chatham House report involve incidents which transpired during the Cold War, and the authors examine how political tensions can affect nuclear decision making.  In one example, a realistic but poorly-timed NATO training exercise in 1983 simulated a nuclear attack and inadvertently put the Soviets on alert. The exercise went ahead in spite of the concerns of the US National Security Advisor, who had recognised that US-Soviet relations were especially bad at the time.[4]

The report finds that regional conflicts also have the potential to escalate quickly and take on a nuclear dimension. Previous conflicts between India and Pakistan, which intensified to the point of nuclear threats, relied on outside mediation to calm tensions. States like Pakistan, where the military wields significant power, might also be more prone to a type of risk-taking that is unpredictable.

Prudent Judgement Saves the Day

Human judgement will always be an imprecise but vital part of nuclear command and control. The alternative – the automation of nuclear weapon launch policies – is fraught with its own profound problems. The Soviets had introduced a semi-autonomous system, ‘Perimeter’, designed to automatically launch nuclear-tipped Intercontinental Ballistic Missiles (ICBMs) at the United States if it detected a launch.[5] The logic behind Perimeter was that in order for a state’s nuclear weapons to deter, the state needs to convince others that it can inflict ‘unacceptable damage’ even if devastated in a first-strike.

Too Close for Comfort raises a deeply disturbing paradox about nuclear weapons. While prudent judgement has saved the day in some instances, miscalculation and misperception have brought us close to inadvertent nuclear use in others. The human factor is therefore a double-edged sword. It is not simply the case that only technical errors can lead to inadvertent use: accurate data still requires decoding and interpretation by fallible human beings so that wrong conclusions are not drawn.


With the current absence of a complete ban on nuclear weapons, the report offers a number of near-term policy options that could potentially reduce the risks of inadvertent nuclear use. These include adopting measures that buy time, such as taking thousands of nuclear weapons off ‘hair trigger’ alert, and retargeting nuclear weapons to the ocean. Recommendations also include increased trust- and confidence-building measures, a wider set of decision-makers involved in nuclear command and control, and educating militaries about the humanitarian impacts should nuclear weapons ever be used again.

One concern emerging from the report is lack of transparency. Nuclear weapons possessors are anxious about revealing details about their nuclear launch policies, and want to avoid embarrassment over instances where they may have come close to launching nuclear weapons due to negligence or miscalculation. There are likely to be other instances where the world has come close to nuclear war by accident, choice, or sloppy practises that we simply do not know about due to secrecy. This poses the question: for how long can the world depend on people making the right calls?  

Sasan Aghlani is a research assistant in the International Security Department of Chatham House, a UK think-tank focused on international affairs. He is a co-author of the report Too Close for Comfort.


[1] Lewis P, Williams H, Pelopidas B, Aghlani S, (2014). Too Close for Comfort: Cases of Near Nuclear Use and Options for Policy. Chatham House Report.

[2] Pp.16-17 of note 1.

[3] Pp.12-13 of note 1.

[4] Pp.13-16 of note 1.

[5] See: Hoffman D (2011). The Dead Hand: Reagan, Gorbachev and the Untold Story of the Cold War Arms Race. Icon Books.