International politics today feels like a collective sleepwalk into the abyss. Three trends are particularly alarming. Artificial intelligence is globalising and turbocharging old and new threats. Meanwhile, no state is strong enough to impose order. And, as in earlier periods of global crisis, geopolitical blocs are hardening, with international cooperation increasingly portrayed as weakness. Alas, there is ample reason to fear a repetition of the worst of history.

Much attention has focused on the risk of war between the US and China, as shifts in hegemonic power historically produce conflict. That is not, however, what is happening. Instead, we are approaching an international order with no hegemon, with several states and regional blocs competing for influence. Using the jargon of international relations theory, we are on the cusp of genuine global multipolarity.

Artificial intelligence is adding further stress to existing power structures. On the one hand, we are all being transformed into godlike figures — soon no innovation or information is beyond our reach. But it is also disempowering everyone, creating new threats and making old ones harder to contain.

Larger states are finding it increasingly difficult to convert power into control. This is already evident in the military domain, where inexpensive drones and AI-assisted targeting have changed traditional hierarchies. The US, with its vast military arsenal, is unable to force an isolated and bankrupt theocracy in Iran to bend the knee; Russia’s failed attempt at subjugating Ukraine is another example of a much more powerful state not getting its way.

New technology is making the loss of control even more evident in other domains. The next pandemic could originate in a hidden laboratory and prove far deadlier than COVID-19. Devastating cyberattacks could be launched from someone’s garage. Frontier AI systems could spiral beyond human control. AI could lower the barriers to developing nuclear weapons, while biological weapons could proliferate without anyone noticing.

Such threats recognise no borders. If one state establishes effective safety measures, nefarious actors will simply move somewhere else and deploy from there. A state can possess overwhelming military and economic capabilities and still be unable to stop a pathogen, prevent a cyberattack or contain the spread of dangerous technologies. Strategies based on peace through strength, regional primacy and zero-sum power dynamics will be like fighting mosquitoes with heavy weaponry — costly and unworkable.

Yet much of the world is doubling down on precisely these strategies. Some state leaders are focusing on winning AI and achieving digital supremacy. Negotiating mutual constraints through international law and institutions is seen as naive or self-defeating. Such approaches are at best ineffective. As global policy, they can destroy humanity. It is as though one can hear the voices of earlier wartime generations: “Do not fall for it again.”And we really shouldn’t, given how unfettered national ambition coupled with weak international rules and institutions has hurt us in the past.

Yet many seem incapable of rising to the occasion. One reason is surely our tendency to underestimate risks that accumulate over time. A 1 percent annual probability of a global catastrophe may sound reassuringly small. Yet, over a century that translates into roughly a 63 percent chance of catastrophe—an intolerable level of risk.

Military strength will remain necessary. But the only way to ensure real security in this new paradigm of multipolarity and artificial intelligence is through powerful international institutions and laws. Global standards must be developed and enforced. International organisations must be allowed to monitor, inspect and report on activities within countries’ borders. Governments will regularly have to share sensitive information. Peace through cooperation must be the new global slogan; not out of altruism but because it is the only way to achieve real security.

Fortunately, we don’t have to start from scratch. Institutions such as the United Nations, the World Trade Organization (WTO) and the World Health Organization (WHO) still retain unparalleled authority and legitimacy. The International Atomic Energy Agency (IAEA) has shown that international monitoring can effectively verify compliance even in highly sensitive areas.

Recent events offer some grounds for optimism. Last year, the WHO adopted a Pandemic Agreement, which will improve the world’s collective readiness to prevent and mitigate future health emergencies. The UN General Assembly recently adopted the BBNJ Agreement to protect biodiversity on the high seas, the first global treaty of consequence in at least a decade. The Security Council has shown that it’s serious about addressing the security challenges posed by AI; during a high-profile debate in September this year, most governments and leading AI developers alike called for stronger global governance and more robust international oversight mechanisms.

These are encouraging signs. But overall, states are investing too little in such global frameworks at a time when the opposite is needed.

Norway has made it a priority to break the curve. We have defined respect for international law and effective multilateral institutions as a core national security interest. We’re investing heavily in UN reform; as one of the UN’s largest financial contributors, we have leverage which we will use to push for greater effectiveness. Alone, Norway is too small to change the course of international politics. But through cross-regional coalitions with like-minded countries, we can make a real difference.

One recent example came in September when Norwegian Prime Minister Jonas Gahr Store and Finnish President Alexander Stubb launched a joint Call for Action to establish global oversight and control over advanced AI models. Countries from all over the globe have joined the initiative. Hopefully, it can serve as a springboard for concrete international action in this field.

No one is in charge in a multipolar world. This new reality is taking shape precisely as technology is making threats harder to contain. The defining challenge of our time will therefore be to couple multipolarity with multilateralism. Building effective institutions when no one can impose order is inherently difficult. But just as previous generations were able to overcome profound challenges, so can we.

It was only after the lived catastrophes of the first and second World Wars that nations united around shared legal principles and institutions. I am convinced that, in time, states will once again realise that this is the only way. What remains to be seen, however, is whether we can muster the necessary resolve without first experiencing another global crisis of world-war-like proportions.