Improving nuclear security

You do not have Javascript enabled. Some elements of this website may not work correctly.

At this very moment, thousands of thermonuclear weapons stand on high alert, ready to be launched in minutes. If used, they could kill tens or even hundreds of millions through fire and radiation. Worse still, these weapons risk causing a "nuclear winter," blocking sunlight for years and potentially killing billions more in the aftermath.

During the Cold War, the US and USSR each built tens of thousands of nuclear weapons. At various points they were nearly used, either accidentally, or deliberately. Fortunately, the end of the Cold War led to major arsenal reductions, but risks still remain. US–Russia and US–China relations have become increasingly strained over the past decade, while India–Pakistan and India–China relations remain tense. While presumably these countries are well aware of the disastrous consequences of nuclear war (deterring them from ever starting one), there is still the possibility of miscommunication, or miscalculation, renewed arms races, and future political or technological developments destabilising the current deterrence regime.

Efforts to reduce nuclear risk mostly require influencing the military and foreign policy establishments in nations with nuclear weapons, encouraging either unilateral or multilateral policies that benefit all involved parties. There are a variety of ways to achieve this, including policy research, political lobbying, track II diplomacy, and public outreach campaigns.

Philanthropists could work to improve societal resilience against the potential chaos and famine that a nuclear war could cause.

Why is reducing the risks from nuclear war important?

When looking at how important a cause is, we can analyse its scale, neglectedness, and tractability. In this case, that means working out:

  • Scale: How bad would a nuclear war be, and how likely is it to occur?
  • Neglectedness: How much work is already being done to reduce nuclear war? The more neglected a cause is, the easier it is for an individual or small community to make a difference.
  • Tractability: How difficult is it to reduce the risk of nuclear war?

What would the scale of a nuclear war be?

To understand the scale of nuclear war as a potential issue, we need to know both:

  • How many deaths a nuclear war would cause.
  • How likely a nuclear war is to occur.

Estimating the number of deaths

A full nuclear exchange between the US and Russia, the nations with the greatest nuclear arsenals on the planet, could kill roughly 50 million people through the direct effects of the blast and fallout. While an India–Pakistan or US–China exchange would see a smaller number of nuclear weapons exchanged, they would still likely lead to tens of millions of deaths from the direct effects. These death tolls would automatically place such wars amongst the worst wars in human history.

Indirectly, a full nuclear exchange would lead to widespread firestorms, which could potentially inject large amounts of soot into the stratosphere. This soot would block sunlight for a number of years, decreasing global temperatures and precipitation in a "nuclear winter."[1] Such an event could potentially kill billions through famine, and, in extreme scenarios, trigger full societal collapse and possibly even lead to human extinction. It should be noted, however, that the likelihood and severity of a nuclear winter scenario is controversial.[2]

Current nuclear arsenals are only a fraction of their Cold War peak (see below). However, technological advancement and economic growth since the Cold War mean that major powers could build up to and beyond their nuclear arsenals quite quickly. Therefore, an arms race between major powers could potentially lead to the deployment of far more nuclear weapons, risking both greater death tolls and an increased chance of nuclear winter.

Estimating the likelihood of nuclear war

Estimating the likelihood of a nuclear exchange is very challenging. Since World War II, no nuclear weapon has ever been used in anger, suggesting that the historical risk cannot have been greater than a couple percent per year. Using information from historical near-misses, one estimate of the historic risk of a US-Russia accidental war was 0.9% per year, but with large uncertainty,[3] while leading national security experts, when polled about the next 25 years, suggested an annual risk of 0.3% for a nuclear conflict killing more than World War II (roughly 80 million people). It's unclear how seriously we should take any of these numbers, but risks of approximately a few tenths of a percent per year seem plausible.

Is reducing the risks from nuclear war neglected?

It's difficult to gauge the neglectedness of this topic. The nuclear powers have strong interests in maintaining control over and protecting their nuclear capabilities — presumably they devote some fraction of the tens of billions of dollars that they spend on nuclear forces to these issues.[4] Attempts to reduce the risk of nuclear war itself, however, must compete with domestic and international political pressures. Militaries, in particular, often face perverse incentives to protect and expand nuclear capabilities.[5]

In 2018, private philanthropic organisations spent around $81 million USD on nuclear security.[6] Of this, $36 million was funded by the MacArthur Foundation. As the foundation recently announced it would be ending its nuclear security grantmaking after 2023, funding will likely be significantly lower going forwards.

It should also be noted that much of the money in this space is devoted to countering nuclear terrorism and preventing proliferation to rogue states, rather than reducing the risks of a full-scale nuclear war — leaving the worst risks as the most neglected.[7]

Is reducing the risks from nuclear war tractable?

Most efforts to reduce the risk of nuclear war require influencing the policy decisions of major nuclear powers, where private philanthropic concerns must compete against the influences of domestic and geopolitical considerations. Work in non-democratic countries such as Russia and China is particularly difficult. In addition, the geopolitical environment can fluctuate dramatically over time, and it's difficult to predict when opportunities will arise.

For example, a series of agreements between the US and Russia in the 1990s led to massive reductions in arsenal sizes as discussed above. In the 2010s, however, relations deteriorated and past agreements were repudiated,[8] driven in particular by Russian aggression in Ukraine and the presidency of Donald Trump.

Whether the next decade will see opportunities to repair relations, and whether private philanthropy has much power to influence this, is difficult to determine. Historically, it appears that private philanthropy has led to at least some reductions in nuclear risks. Probably the clearest case was the role played by philanthropically funded research in the origins of the Nunn-Lugar agreement, while the New START treaty and the JCPAO provide other plausible examples.[9]

How can we reduce the risks from nuclear war?

As already noted, reducing the risks from nuclear war usually requires us to influence the policies of the major nuclear powers. There are many specific policies one could push for, but they are often controversial. Nevertheless, some plausible goals include:

  • Reducing arsenal sizes.
  • Removing particular destabilising weapons (or preventing their construction), such as nuclear-armed cruise missiles.
  • Committing to "No first use" policies.
  • Committing to not target communications networks, cities, or nuclear power stations.
  • Preventing proliferation of nuclear weapons or materials to additional countries.
  • Reducing stockpiles of fissile material.
  • Improving relations between nuclear powers.

It's difficult to know which goals are best, or even whether pursuing a given goal is good on balance. For instance, it might seem obvious that nuclear arsenal reductions are good: the fewer nuclear weapons in the world, the fewer that may be launched in a nuclear war. However, reducing arsenal sizes could plausibly increase risks of war, precisely because the severity of a war has decreased. Likewise, if efforts to reduce arsenals are primarily focused (or primarily effective) in the US but not China or Russia, then this might increase the strength of authoritarian regimes at the expense of more liberal ones. The point of this example is not to suggest that reducing nuclear arsenals is a bad goal, all things considered, but to highlight that pursuing even seemingly good goals could have consequences that are difficult to predict.

To work towards any of these goals, philanthropists could target a specific country (usually the US), or aim to foster agreements between countries. Such objectives could be achieved through technical policy research, political lobbying, public advocacy, and track II diplomacy.

Another approach is to work to reduce the effects of worst-case nuclear wars. For example, we could improve societal resistance to massive food shocks (such as those produced by nuclear winter) by investing in the research and development of alternative food sources.

Why might you not prioritise reducing the risk of nuclear war?

There are more serious risks

Full-scale nuclear war appears unlikely. Such an exchange between two major powers would be undoubtedly disastrous for both countries, so they have strong incentives to avoid such a catastrophe — this is illustrated by the fact that no nuclear weapon has been used in anger since World War II. Furthermore, even in the event of a full-scale nuclear war, human extinction appears unlikely.[10]

In addition, nuclear weapons technology is mature and well-understood, so the scope for new and unexpected risks in this space is relatively low. You may want to compare the risks from nuclear weapons with those from emerging technologies, particularly artificial intelligence and biotechnology. Unlike nuclear weapons, these technologies are rapidly evolving, and the pathways for catastrophe or extinction are more numerous and less well-understood. It therefore appears that they present considerably higher risks than nuclear weapons.[11]

There are more tractable problems

Influencing the policy of major nuclear powers is incredibly challenging, and the tractability of these approaches varies greatly as the geopolitical landscape shifts. Furthermore, even if there are major risk reductions, it would be difficult to attribute these to the impact of philanthropists, as opposed to other changes in the world.

For these reasons, you may wish to work on problems where there are more straightforward paths to impact.

What are some charities, organisations, and funds trying to reduce nuclear risk?

Disclaimer: We have not yet been able to investigate which organisations we recommend supporting if you want to reduce or mitigate the risk of nuclear war. Below are some organisations we know are working in this space, but their inclusion is not an endorsement.

Learn more

We recommend the following resources to learn more about nuclear security:

We also recommend the following books, but note that these are mostly focused on the history of nuclear weapons, so their relevance to current nuclear security issues is not always that clear:

Our research

This page was written by Damon Binder. You can read our research notes to learn more about the work that went into this page.

Your feedback

Please help us improve our work — let us know what you thought of this page and suggest improvements using our content feedback form.


  1. The original paper is here, following earlier work pointing out that the smoke after a nuclear war could have significant climate effects. A more recent review can be found here. ↩︎

  2. See here for a discussion of the early debates about nuclear winter, including the alleged political motivations of the original scientists. There is also more recent modelling in favour of and against the hypothesis. ↩︎

  3. Their 90% confidence interval was 0.02%–7%, which means they were 90% sure the probability per year falls between 0.02% and 7%. ↩︎

  4. The US, which has the largest budget for nuclear forces, is projected to spend somewhere around $60 billion on their nuclear forces per year over the next decade. ↩︎

  5. This is discussed by Daniel Ellsberg in his book The Doomsday Machine: Confessions of a Nuclear War Planner. Ellsberg also discussed this topic, as well as how corporate interests played a role, on The 80,000 Hours Podcast. ↩︎

  6. The numbers are taken from page 28 of this report by the MacArthur Foundation. ↩︎

  7. For instance, the Nuclear Threat Initiative (which received $21 million in 2018, making it by far the largest recipient of grants) has historically focused primarily on nuclear terrorism and secondarily on preventing profileration, as described in this summary of a 2014 interview with GiveWell. ↩︎

  8. In particular, the US formally withdrew from the INF in 2019 and the Open Skies Treaty in 2020. ↩︎

  9. See Philanthropy, Nuclear Nonproliferation, and Threat Reduction , by P. Rubinson (2021) for a more detailed discussion. ↩︎

  10. The issue of whether we could rebuild after a global catastrophe, such as a nuclear war, was discussed on The 80,000 Hours Podcast. ↩︎

  11. See, for instance, Chapter 6 of The Precipice by our co-founder, Toby Ord. ↩︎