How To Prevent Weapons of Mass Destruction in a Changing Geopolitical and Military Landscape
Global Dispatches is partnering with the Global Challenges Foundation for a series of episodes examining how to guard against certain events or scenarios that, should they unfold, would threaten large swaths of humanity across multiple continents. These are often referred to as Global Catastrophic Risks for their potential to disrupt lives and livelihoods on a massive — even planetary — scale. The Foundation is dedicated to raising awareness of global catastrophic risks and strengthening global governance to address them. Global Challenges Foundation’s 2026 Global Catastrophic Risks report outlines five of the biggest risks facing humanity today, including from Weapons of Mass Destruction, the topic of this episode. You can find this report here.
Two of the authors of the chapter on weapons of mass destruction are my interview guests today. Wilfred Wan is director of the Weapons of Mass Destruction Programme at the Stockholm International Peace Research Institute (SIPRI) and a researcher in earth system governance at the University of Oslo. Michael Wernstedt is head of Common Security at the Global Challenges Foundation.
We kick off by discussing how geopolitical and military trends are increasing the risks surrounding weapons of mass destruction, before turning to a longer conversation about how to strengthen international cooperation and global governance to prevent the use of WMDs—and the catastrophe that would entail.
The episode is freely available across all major podcast platforms, including Apple Podcasts and Spotify. You can listen directly below.
Transcript edited for clarity
Michael Wernstedt: We know from history that sometimes there can be leapfrogs and windows of opportunity, and then it’s vital to have the solutions ready, then we can’t start working on the solutions.
Mark Leon Goldberg: Welcome to Global Dispatches, a podcast for the foreign policy and global development communities and anyone who wants a deeper understanding of what is driving events in the world today. I’m your host, Mark Leon Goldberg. I am a veteran international affairs journalist and the editor of UN Dispatch.
Enjoy the show.
Today’s episode is produced in partnership with the Global Challenges Foundation. The Foundation is dedicated to raising awareness of global catastrophic risks and strengthening global governance to address them. Global Challenges Foundation’s 2026 Global Catastrophic Risks Report outlines five of the biggest risks facing humanity today, including from weapons of mass destruction, the topic of this episode. You can find this report at globalchallenges.org/GCR-2026.
Two of the authors of the chapter on weapons of mass destruction are my guests today. Wilfred Wan is director of the Weapons of Mass Destruction Programme at the Stockholm International Peace Research Institute, SIPRI. Michael Wernstedt is the Head of Common Security at Global Challenges Foundation. We kick off by discussing how both geopolitical and military trends are increasing the risks around weapons of mass destruction, before turning to a longer conversation about how to strengthen international cooperation and global governance to prevent the use of WMDs and the catastrophe it would entail.
To access other episodes in this series, please visit globaldispatches.org. Now here is my conversation with Wilfred Juan and Michael Wernstedt.
So, we’re here to discuss the catastrophic risk posed by weapons of mass destruction. But WMDs have not just been a catastrophic risk, but really an existential threat for decades. I’m curious to learn from you both what you think distinguishes the dangers of our current geopolitical landscape from other seemingly perilous moments of the nuclear era.
Michael, why don’t you go first?
Michael Wernstedt: If we just zoom out to the geopolitical landscape, I think it’s clear that the 30 years after the fall of the Berlin Wall has proven to be an exception in history. I grew up in that era, and then I grew up with the narrative that as prosperity would grow, democracy would spread and wars would be a thing of the past. And since a few years, it’s obvious that that is not the case. It was enabled by us having one superpower, among other things. But now we see geopolitical fragmentation and rising geopolitical tensions. And parallel with that, we also see that globalization has not fulfilled the promise of growing prosperity for all, which is called the rise of populism and authoritarian leaders.
And both these things have put a higher pressure and focus on nation states and national security, which has contributed to increased conflicts around the world. And then we can add climate to that, where food scarcity and land scarcity have also contributed to increasing conflicts. And when nations have been threatened, they have more so than in the past feel inclined to abandon the norms that have served us since the Second World War. So, that is the backdrop. And of course, it’s not an extensive list, but it’s a few things at play. And then if we hone in on the nuclear era, in February, the New START agreement expired.
Mark Leon Goldberg: And this was the agreement between Russia and the United States to limit their stockpiles of nuclear weapons. It was the kind of last remaining bilateral nuclear agreement between the United States and Russia.
Michael Wernstedt: Exactly. And that is a successor to previous agreements. And there’s really been agreements for the last 50 years. This is the first time in 50 years that we don’t have that kind of agreement. And the stockpile has been reduced for the last 40 years. But last year was the first time that actually started increasing again. So that is really a shift that we haven’t seen in 40 years. And the second aspect is that the nuclear doctrine has started to change. So historically, we’ve had a commitment to no-first-use. But a number of countries have changed their doctrine, and especially Russia have changed their doctrine, and really played with a nuclear taboo — the taboo that we won’t use nuclear weapon because we know there are no winners if we do.
Mark Leon Goldberg: And Wilfred, it seems that layered on top of this profoundly different geopolitical landscape that Michael just articulated is the changing nature of warfare itself, which seemingly would add an additional risk. And in the Global Challenges Report, you articulated and defined this idea of multi-domain escalation risk. What do you mean by that?
Wilfred Wan: Sure. Let me start by talking about escalation in general, which in this context, means conventional warfare that can escalate to the use of nuclear weapons, whether deliberate or inadvertent. Now, that’s not a new possibility. That was a central concern during the Cold War. But what I would argue is that this possibility is increasing, and that’s because there’s more varied pathways to escalation. And so that’s where the multi-domain of it comes in. So, when we talk about domains, we generally refer to kind of the traditional military operational domains of land, sea, and air. And now we’re including cyber and outer space, as well as information warfare into that.
There’s not necessarily a consensus in terms of kind of the declaration of these as military or operational domains. It’s not ubiquitous or common across states. But multi-domain thinking is just increasingly a part of military operations, of planning, and of some of the doctrines that Michael referred to. So basically, across a number of nuclear-armed states, there’s acknowledgement that conflict will likely unfold across these multiple domains. We are already seeing this in recent and ongoing conflicts in Ukraine, in Iran, with the use of AI, of artificial intelligence, of drones, cyber operations, attacks on space capabilities, all in a coordinated manner.
But this kind of conflict or these kind of operations really complicates the nuclear landscape for a number of reasons. One is that you have more relevant capabilities from a strategic perspective, more relevant capabilities and operations, including capabilities that can make nuclear forces vulnerable.
Mark Leon Goldberg: What would be some of those capabilities that might make a nuclear force vulnerable? And these are capabilities, presumably not like a first strike against another nuclear force with nuclear weapons, but something in one of these other domains?
Wilfred Wan: Yes, exactly. So, you’re talking about things like advanced conventional weapons, missiles, and things like this, where if you have a combination of cyber operations that attacks the adversary’s early warning systems, it makes these missiles capable of performing the same kind of missions that used to be preserved for nuclear weapons. So, a second point is that that you have more states who can expose those vulnerabilities, including non-nuclear armed states, because they possess and deploy these new capabilities, including in terms of cyber operations. So, last June, we saw a number of Ukrainian drones target and damage Russian strategic bombers because these bombers were being used to launch conventional cruise missiles into Ukraine.
But by doing this, they also attacked Russian nuclear forces, which were undermined as a result of the attacks. And in fact, some Russian experts called for a nuclear response to those attacks because of the strategic impact that they had. So, a third point and related is just the entanglement of capabilities in new domains. And what I mean by entanglement is that a lot of these capabilities in space and cyber, they can serve conventional and nuclear missions. They can serve civilian and military missions in a manner that can contribute to confusion or unintended consequences.
So, none of that necessarily means that there will be a nuclear use or nuclear response to these things. But basically, multi-domain operations contribute to greater destabilization and more frequent crises because there’s a lack of agreements on thresholds, on red lines, what would be a proportionate response, and because there’s potential for spillover effects in the context of alliances and strategic partnerships. And all of that, in turn, then impacts on escalation pathways.
Mark Leon Goldberg: So, Michael, it would seem that we have both this changed geopolitical landscape, a changed warfare landscape in terms of the ability of nuclear forces to be threatened by nonconventional means and the multi-domain escalatory risk that Wilfred just articulated. Surveying the world today, are there near-term threats that you see where these two trends might collide?
Michael Wernstedt: One aspect that we haven’t touched upon so much is also the development of AI in the military domain and particularly decision support systems. And there we see a rapid development. And I think the big risk that we are all worried about is that an AI system would suggest or even have the autonomy itself to respond to a supposed threat and launch a nuclear attack. And that could, of course, cause unexpected consequences. I think humans still have that sort of ingrained taboo in them that I don’t want to be the first one to start a nuclear war. But there is not a necessity that an AI system would have that.
Mark Leon Goldberg: Because nuclear weapons have been around for so long and because of their existential threat to humanity, there have been attempts, over the decades, to create like an architecture of global governance to control their use, to encourage disarmament. I’m thinking principally right now of the Non-Proliferation Treaty, and we’re speaking about a month ahead of the review conference for the Non-Proliferation Treaty, which happens every five years, which is an opportunity for member states to gather and discuss ways to strengthen non-proliferation regimes and nuclear security.
But I’m curious to learn from you first, Michael, and then Wilfred, whether these current global architecture or whether this current system of global governance is sufficient to address the complexity of nuclear security today.
Michael Wernstedt: I think it’s helpful in this aspect to look at Bill Sharpe’s Three Horizons network, to look at the first horizon with the current government system and see what do we want to preserve there. And then look at the third horizon, which is emerging, to see what do we want to envision in the future. And then look at the second horizon to see what could bridge the gap between the two. In the first horizon, the NPT, the Non-Proliferation Treaty, is really a cornerstone. It doesn’t have an expiration date, which is lucky. So, it will continue to exist, even if the review conference is not successful.
However, the legitimacy of the treaty is really at risk if the review conference is not successful, because if the review conference fails, states will be less inclined to respect the treaty. And the last two review conferences have failed, which is unprecedented. So, if the third one failed, and unfortunately it looks a bit bleak given the geopolitical landscape, that will really be risky. And then it is vital that we find the successor to the New START agreement and continue the arms reduction. And then I would also like to add the accountability aspects.
We have the international courts, the International Court of Justice that settles conflicts between state and the International Criminal Court that prosecutes individuals that are charged for crimes against humanity. And it’s vital to protect their integrity and jurisdiction, especially the International Criminal Court has been under massive threat, especially from the U.S. Then if I look at the second horizon, which is slightly further, what can we build? We need a treaty on the use of AI in the military domain. There is a declaration, and 45 countries have signed that declaration, but there is no accountability mechanisms. And of course, there’s no universality of it.
And then we need governance for the mechanisms of multi-domain escalatory risks. And there, I think Wilfred could add more insight into what that would look like. And then there is the TPNW, the Treaty on the Prohibition of Nuclear Weapons. So, the idea of a nuclear weapon-free world. And it might sound utopian, but 95 countries have signed it and 74 ratified it. So, it is there and it is possible. So that’s the content side. And then on the structure side, if I paint the vision, I think we need the global decision making body that could be inspired maybe how the EU works, that we can make decisions by majority decisions.
So, not that one country can block the whole process. But if a majority is for, probably a qualified majority, the legislation can move ahead. And I think in order for that to happen, there needs to be some kind of legitimacy. So, there needs to be probably a represented body with direct representatives, a bit like the EU parliament functions. And this might sound utopian, but I’ve been working with social change for the last 15 years, and I think there are several paths that could lead us there.
I like to look at Everett Rogers’ theory of diffusion of innovation, which is how really innovation can spread in society, but also how ideas can spread in society, where you look at different groups, first the innovators and the early adopters, early majority, late majority, and then the laggers. And rather than trying to convince the laggers, you just look at what group is convinced now, and how can we convince the next group? One example how this is used in another realm is the LBGTQ community in the U.S., where they said, “Okay, let’s stop fight the conservative Christian groups, and let’s instead try to just convince the next group.”
And within two years, they had managed to get same-sex marriage through in 18 states. And of course, we see both on the nuclear side that 95 states have signed. So, we actually have an early majority that is on board already. And on the AI side, we have 45 states that have signed. So, there we have early majorities on board. And then it’s just a question, how do we get the late majority on board? And also, I think another actor that is, or group of actors that is interesting to look at is actually AI company. I think it’s interesting what Anthropic did in relation to Pentagon, where they said, “No, we need guardrails on how our software, our AI is used.” And that could also be a force to be reckoned with if that is an industry-wide initiative.
This is an ongoing slow process, but we know from history that sometimes there can be leapfrogs and windows of opportunity. And then it’s vital to have the solutions ready, then we can’t start working on the solutions. So, I think a bit or a lot of our work is to have the solutions ready. And we have the solutions for the reform of the UN in the Summit of the Future. We have the TPNW, the Treaty on Prohibition of Nuclear Weapons. We have emerging legislation on the AI. What we don’t have yet is the legislation on multi-domain escalatory risks. And that’s why I think Wilfred’s work is so interesting.
Mark Leon Goldberg: And on that point, Wilfred, what opportunities do you see for systems of international cooperation or global governance to help control or manage the multi-domain escalation risk that you articulated?
Wilfred Wan: Well, let me start by looking more broadly at nuclear governance. And you asked this question of whether the existing global governance structure is sufficient. I mean, it is if it’s being adhered to, if it’s being strengthened and it’s being implemented. Both of you refer to a New START at the beginning of this conversation and the fact that it has expired. But it’s also linked, unfortunately, to kind of a trend over the last few decades in terms of the deterioration of the arms control architecture with withdrawals from the Anti-Ballistic Missile Treaty, the Intermediate Nuclear Forces Treaty.
You look around, there’s discussions about potential resumption of nuclear testing. And compounding the issue as just the greater value placed on nuclear weapons. All of the nuclear armed states are undertaking modernization programs. A number of non-nuclear armed states are leaning more into reliance on nuclear deterrence or having conversations about nuclear sharing. So, we’re heading towards a world where there will likely be more nuclear weapons, less cooperation, let alone information about them. So, when we talk about the NPT, it really rests on this bargain between nuclear weapon states and non-nuclear weapon states.
That requires disarmament progress, and we’re not seeing that. So, I think it’s really important to highlight this because that is fundamentally the most effective way of addressing multi-domain escalation risk. At the same time, I do think, while we recognize the difficulty of the strategic environment of the nuclear landscape, there are some near-term opportunities to enhance governance on these issues. There’s value even in like-minded states, for instance, in alliance relationships, exchanging views on threat perceptions, on risk assessment, across multi-domain interactions, to basically establish a common value structure, considering what states see as the roles and strategic valuations of capabilities, or to identify the activities or behaviors that they find especially escalatory, and how the inclusion of cyber and outer space in particular interact with traditional domains of warfare from the perspective of nuclear deterrence.
This kind of exchange is necessary to have future agreements on these kind of issues. In the nuclear risk reduction conversation, there’s a lot of conversation about conflict prevention and conflict management toolkits. So, what I mean by that is the establishment of crisis communication channels, pre-launch notifications, establishment of designated points of contact. And these can be updated more systematically to account for multi-domain interactions. For instance, by covering AI issues, cyber issues, space issues.
Mark Leon Goldberg: Have like a red phone sort of thing for like AI or cyber issues so that the parties have the opportunity to speak directly to each other to reduce any confusion over some unknown event. Kind of like you have in like the nuclear space, at least between Russia and the United States, less so between the United States and China, although that’s trying to be developed and established.
Wilfred Wan: Precisely, to kind of establish these communication channels to prevent misunderstanding, misinterpretation. Now, some of these channels across certain countries exist, but it’s quite inconsistent, as you flag, to increase military-to-military dialogue on these issues as well. But when we talk about kind of more longer-term aspects, there needs to be a more holistic approach to nuclear risk. And this requires a more forward-looking approach to dealing with strategic technologies. So, to give you an example, scientific and technological developments are being considered by working groups by permanent bodies in the Chemical Weapons Convention, in the Biological Weapons Convention, and in the Nuclear Ban Treaty, the TPNW that Michael referred to.
But it’s not considered in the NPT, which is the nearly universal nuclear treaty. Developments across domains, across capabilities, fundamentally are changing how nuclear weapons are being thought about and when they’re being considered for use. And so these developments absolutely need to be considered, whether through a regular review of these developments, these technological advances, through more interactions across these different UN processes that are quite siloed in how they tackle these issues, through more inclusive engagement with the private sector and industry, and Michael gave the great example of anthropic.
Or just simply more explicit discussion about multi-domain risk. It often feels like military thinking around newer domains, how best to use related capabilities to achieve battlefield victories, to bolster deterrence, etc., tends to outpace thinking about how these same actions could be misperceived or reacted to or what accompanying risk is. And so, in this context, the push for first mover advantage, that is states wanting to take advantage of these technologies to win, can also mean that there is an unearned confidence in our ability to control or manage escalation. So, I think the point that I want to end with here is just a need to be a bit more humble in how we think about these things and recognize the uncertainties that exist.
Mark Leon Goldberg: So, I’d like to conclude by learning from both of you what potential near-term opportunities you see in the coming months or years to implement some of these ideas around governing either multi-domain risk or weapons of mass destruction more broadly. Are there some like kind of diplomatic moments coming up where some of these ideas might be productively advanced?
Michael Wernstedt: I mean, the first clear opportunity is in the next month with the NPT review conference. So, that is a real clear opportunity. And then there are a number of initiatives on the AI domain. So we have the responsible AI in the military domain initiative that we are also in discussion with, which I see is one initiative. And then I think, as I mentioned, that there could also be private initiatives with Anthropic leading the example.
Wilfred Wan: I would also start with the NPT Review Conference. And while AI issues have been discussed briefly in that context, it’s been primarily limited to integration of AI in nuclear command control and communication systems. And yet there are so many other ways in which AI can contribute to thinking around nuclear weapons. And so, I think the Review Conference is an opportunity to really expand that conversation about AI, but also to look at some of these other domains and activities around them. The second opportunity I would point to is the convening of the open-ended working group on space security issues. And the next meeting will take place in July of this year.
So, to, again, connect outer space issues to strategic stability and broader issues concerning nuclear risk, I think, is really important, again, to my point about de-siloing conversations. The last thing I would say that hasn’t come up so much in our conversation so far, but I want to highlight, is the need to discuss these things at a regional and sub-regional contexts. Because ultimately, when you talk about WMD, when you talk about valuation of nuclear weapons, and so forth, these issues don’t exist in a vacuum, and they’re linked to underlying regional, sub-regional security concerns and sub-regional and security environments. And so, to explore how these capabilities and these operations take form in those settings can be a concrete way to move forward.
Mark Leon Goldberg: Michael and Wilfred, thank you so much for your time. I really appreciated your contributions to this report. Thank you.
Michael Wernstedt: Thank you so much.
Wilfred Wan: Thank you so much.
Mark Leon Goldberg: Thanks for listening to Global Dispatches. The show is produced by me, Mark Leon Goldberg. It is edited and mixed by Levi Sharpe. If you are listening on Apple Podcasts, make sure to follow the show and enable automatic downloads to get new episodes as soon as they’re released. On Spotify, tap the bell icon to get a notification when we publish new episodes. And of course, please visit globaldispatches.org to get on our free mailing list, get in touch with me, and access our full archive. Thank you.