Isidora avatarIsidora & Cos Explain

Russia Fired Cluster Warheads at Kyiv. Why Do 112 Countries Ban Them While the Biggest Armies Don't?

A misty rice field in a river valley at sunrise, with red-and-white tape and a small red flag marking a rusty round object half-buried in the soil
Summary

On 22 October Russia hit central Kyiv with ballistic missiles carrying cluster warheads. A 2008 treaty bans these weapons in 112 countries, yet Russia, Ukraine, the US, China, India, Israel and several of Russia's European neighbours stay outside it. What cluster munitions are, why they kill long after a war ends, why so many armies still want them, and whether the ban is holding. Analysis with a short opinion section.

On the night of 21 to 22 October, three of the ballistic missiles Russia fired at Kyiv carried cluster warheads, according to President Zelensky. One burst over Pechersk, the historic centre, between the Kyiv-Pechersk Lavra and a monastery. A Ukrainian adviser identified it as an Iskander with a warhead that scatters 58 bomblets over roughly 400 by 400 metres, Euronews reported. Counts of the dead from the whole night differ: at least two in Kyiv according to Zelensky, at least four nationwide according to Ukraine’s foreign ministry, cited by CBS.

Most of the world has banned this kind of weapon. Russia has not. Neither have Ukraine, the United States, China or India. Why?

What a cluster munition is

A cluster munition is a container, a bomb, shell, rocket or missile warhead, that opens in the air and releases dozens or hundreds of smaller explosives over an area. One round covers ground that would otherwise need many. That is the problem, twice over. Over a city, the spread cannot tell a soldier from a café. And a share of the bomblets fail to explode, then lie in fields, roofs and gardens like small landmines, waiting for a farmer or a curious child.

Laos and Lebanon: why the ban came

The clearest case is Laos. Between 1964 and 1973 the United States dropped more than 270 million submunitions on the country; the Lao clearance agency estimates that about 30% did not explode, leaving roughly 80 million behind.

The turning point came in 2006. In the last days of the war in Lebanon, Israel fired an estimated four million submunitions into the south, about 90% of them in the final 72 hours, Human Rights Watch documented. Frustrated that the UN’s weapons forum ran on consensus, Norway and a group of middle powers and affected states started a separate negotiation. The resulting Convention on Cluster Munitions was adopted in Dublin in May 2008 and entered into force in 2010. It bans use, production, transfer and stockpiling, with a narrow exception for weapons with fewer than ten heavy, sensor-guided, self-destructing submunitions.

What the ban has achieved

The Cluster Munition Monitor 2026, published by the campaign coalition that pushed for the treaty, counts 112 member states. It reports no confirmed use, production or transfer by any member since joining, and the destruction of 1.49 million cluster munitions holding 179 million submunitions. The stigma also reaches outsiders: the 2025 edition notes that Israel and Cambodia, neither of them members, protested when such weapons were allegedly used against them.

But the human toll is rising. The Monitor recorded at least 1,063 casualties in 2025, one of the highest totals it has ever recorded, 87% of them civilians, and 927 of them in Ukraine. As an advocacy body counting only verified cases, its figures are minimums. In 2025 and 2026 it found use by Russia, Ukraine, Myanmar, Thailand (against Cambodia) and Iran (against Israel), all outside the treaty. It found new production in nine countries, including China, India, Israel, North Korea, Poland, Romania, Russia and South Korea, and a US$211 million contract for Israel to make cluster munitions for the US, according to its press briefing.

Why the big armies stay out

The holdouts’ reasons are not all cynical.

The military argument. Against troops in trenches, a cluster shell is far more efficient than a single explosive one. Writing for RUSI, a British defence think tank, Jack Watling and Justin Bronk cited US Army Vietnam data: 13.6 ordinary 155mm rounds per enemy soldier killed, against 1.7 cluster rounds. For an army short of shells and barrels, that matters enormously.

Ukraine. Kyiv asked for American cluster shells in 2023 to use on its own occupied land, which is already full of Russian mines, and gave written assurances about careful use and later clearance. Washington sent them as a “bridge” until shell production caught up, saying their failure rate was under 2.35% against 30 to 40% for Russia’s. US law itself caps exports at a 1% failure rate, which the decision bypassed, the BBC noted. Ukrainian use has also killed civilians: Human Rights Watch documented at least eight deaths from Ukrainian cluster rockets in occupied Izium in 2022.

Russia. Moscow says the weapons are lawful when used properly. In 2023 Putin warned that Russia had a “sufficient stockpile” and reserved the right to respond in kind, although Russian use had been documented since 2022. Russia was the only country to vote against the 2024 UN resolution supporting the treaty, per the Monitor.

China and the “regulate, don’t ban” camp. China says it has never used cluster munitions abroad, that irresponsible use and transfer are the real problem, and that the UN weapons forum, which includes all major producers, is the right place to act; it also argues that whoever uses them should clear them, in a statement to the treaty’s members. That alternative was tried. In 2011 a weaker protocol backed by the US, Russia, China, India and Israel, which would have banned only older types, collapsed when more than 50 states refused it, the Arms Control Association reported.

Russia’s neighbours. Finland never joined, citing its long border with Russia and the cost and time of replacing its stocks; in 2024 it said conditions for joining had not arrived “considering our geographical location”, according to the Monitor. From 2015 to 2018, Poland, Estonia, Greece and Romania issued joint UN statements with it about their “legitimate security concerns”. Lithuania went further: it joined in 2008, then became the first country ever to leave, effective March 2025, arguing in its withdrawal notice that modern weapons with self-destruct fuzes reduce civilian risk. The treaty’s presidency called the move one that risks eroding the norm. The Baltic states, Finland and Poland have since left the landmine ban treaty too, Canada’s parliamentary library notes.

Do “safer” cluster munitions fix the problem?

This is the crux, because both the US and Lithuania lean on it. The evidence is mixed. Test-range failure rates are measured in good conditions; battlefields have mud, sand, trees and roofs. After 2006, Norwegian People’s Aid studied M85 submunitions in Lebanon, a type fitted with self-destruct fuzes and advertised as failing about 1% of the time or less. On the ground it found failure rates of around 10%. Independent combat data on newer designs is thin.

Opinion: a ban that works on those who signed it

The treaty has done what treaties can do. Its members have stopped using these weapons, destroyed huge stockpiles and cleared land in places like Laos and Lebanon. Affected states in Asia, Africa and Latin America led that as much as Europe did.

But I don’t think it’s fair to call every holdout a hypocrite. A country that expects to fight Russian armour on its own soil faces a real choice between more shells it may not have and fewer, deadlier ones that leave a mess it will have to clear itself. Ukraine made that choice under invasion. I understand it, and the moral weight is not equal: firing at a monastery district in a capital is not the same as firing at trenches in occupied fields.

Still, “safer” is the claim that worries me most. The Lebanon data shows how optimistic factory numbers can be, and the children who find bomblets years later never chose any of it. If the norm is to survive, the realistic steps are modest: no use in populated areas by anyone, published failure-rate data from actual battlefields, and money for clearance and survivors, which US aid cuts have just reduced, closing Laos’s main victim-assistance project. A weapon that keeps killing after peace is signed deserves more scrutiny, not less, from the states that still keep it.

Sources

Comments