Lloyd's Retreat from Hybrid War Risk

II.Hybrid War, or the Dissolution of the Categories

A war that stays below the threshold

Hybrid war, sometimes called gray-zone conflict, covers operations that stay short of open conflict but still impose heavy costs and uncertainty on the adversary. It is waged through cyber-attacks, sabotage of infrastructure, jamming of navigation signals, disinformation, by any means that wounds without crossing the formal threshold of war. Since the start of 2024, at least eleven undersea communication and power cables have been damaged in the Baltic Sea, and informal NATO sources point to a significant share of transatlantic cables affected by suspicious causes². The ships implicated often belong to a shadow fleet under flags of convenience, and dragged their anchors along the seabed. The strategy is deliberately ambiguous, cheap to run and highly rewarding.

The four properties erased

For insurers the problem is simple: hybrid war erases, one by one, the four properties on which the exclusion rested. It is not declared, since it stays below the threshold of assumed conflict. It is not attributable. Its authors blur responsibility by working through intermediaries, such as ships under third-country flags or criminal and diaspora groups. It is not bounded, for a cyber operation spreads without borders and a cable sabotage produces chain effects far beyond its target. It is not kinetic, often leaving only a line of code or an anchor that one may say was dragged by accident. The episode of the ship seized and then released by Sweden in early 2025, for want of being able to prove intent, sums up this indeterminacy³. The exclusion turns on one question, does this loss arise from war, and it no longer has a certain answer.

From crime to war, a continuum

A subtler confusion sits on top: the line between crime, espionage and war is blurring. Ransomware offers the clearest example of this murky zone. An attack may be carried out by a criminal group acting for profit, but tolerated, sheltered or steered by a state that finds a strategic interest in it. Is that a criminal act, covered by the cyber policy, or a state operation that an exclusion carves out? In reality there is no clean line, only a continuum, where each attack mixes profit motive and state interest in varying degrees. Yet the exclusion assumes a binary cut. Either the attack is state-backed or it is not, when the phenomenon is gradual by nature. This mismatch between the binary logic of the clause and the continuous nature of the threat is one of the deepest sources of coverage uncertainty.

III.NotPetya, the Precedent That Triggered Everything

A weapon of war that overflows the battlefield

This breakdown has real costs. It has already run into the billions and nearly destabilized the market. On 27 June 2017, the eve of Ukraine's national holiday, a piece of malware called NotPetya got in through an update of Ukrainian accounting software. Within hours it had spread to more than sixty countries, and it caused around ten billion dollars of damage worldwide⁴. The United States attributed it in 2018 to Russian military intelligence, and US justice indicted six GRU officers in 2020. Merck, the pharma group, lost around 1.4 billion dollars. Mondelez lost more than one hundred million. Maersk, the carrier that handles nearly a fifth of world trade, lost between two hundred and fifty and three hundred million. A weapon designed to strike Ukraine had overflowed onto the world economy, exposing insurers who had never priced such a risk.

Maersk shows what that overflow looks like. Within hours, almost all of the shipping group's computer systems went dark worldwide. Its port terminals stopped, and a noticeable share of global maritime trade was disrupted, even though the company had nothing to do with the conflict in Ukraine. That uncontrolled spread is the signature of cyber war risk. Unlike a missile, a cyber weapon has no bounded target and moves blindly through interconnected software chains. The war loss, once confined to a theater of operations, thus becomes planetary, and it is this absence of boundary that forbids any classical pooling.

The ruling that sounded the alarm

Merck claimed under its all-risks property policies, which carried 1.75 billion dollars of coverage. Its eight insurers, Lloyd's syndicates among them, refused to pay nearly 700 million, citing the hostile or warlike acts exclusion⁵. In January 2022, the Superior Court of New Jersey ruled against them. The exclusion had been drafted for classical armed conflict, the court held, so it could not reach a cyber-attack, and the insured could not reasonably have expected it to carve out such a loss. The appellate court upheld that reasoning in 2023. It found NotPetya to be a non-military cyber-attack on an accounting software provider, not a hostile or warlike act within the meaning of the clause. The insurers finally settled in early 2024 to avoid an unfavourable precedent imposing itself on the whole market.

IV.Lloyd's Response, Redrawing the Frontier

The exclusion of state-backed cyber-attacks

The response of the London market matched the alarm. Through its market bulletin Y5381 of 16 August 2022, Lloyd's required that, from 31 March 2023, every standalone cyber policy contain a clause excluding losses resulting from a state-backed cyber-attack⁶. Where there was no separate war exclusion, the clause had to carve out losses from war, declared or not. It also had to exclude state-backed cyber-attacks that could significantly impair the functioning of a state or its security capabilities, and it had to set out in advance how the parties would establish that an attack came from a state. Behind the directive lay the fear of systemic accumulation.

The dread of the systemic loss

The figure that haunts underwriters is known. A model built by Lloyd's with the Cambridge Centre for Risk Studies puts the cost of a cyber-attack on a major payments system at up to 3.5 trillion dollars over five years for the world economy, including 1.1 trillion for the United States alone. The entire global cyber insurance market wrote only around 9 billion dollars in premiums in 2022, by the same study's count⁹. Set a potential loss in the thousands of billions against a few billion of insurance capacity, and the retreat makes sense. To frame the risk, the Lloyd's Market Association published a series of model clauses, whose gradation reveals the difficulty of the exercise⁷.

The gradation shows that Lloyd's did not try to exclude every state-backed cyber-attack. The broadest clause would allow that, but the market does not use it. The target is the most devastating attacks, the ones that hit a state's ability to function. A later bulletin, Y5433, refined the framework in 2024 without loosening it. It extended the scope to multi-line policies, and it required a distinct, explicit grant of coverage for any cyber-attack carried out within a conventional war⁸. The market thus redrew, clause by clause, the frontier of what it agrees to bear.

This is the market's second front, after silent cyber. The London market had first sought, through the so-called affirmative cyber clauses, to drive cyber risk out of the traditional policies that covered it without naming it, a subject already treated in this series¹². The state-backed war exclusion goes a step further and targets the most catastrophic fringe of cyber, the attacks of state origin. In both cases, the approach is the same, to turn an implicit, unpriced exposure into explicit, bounded cover, even at the price of carving out the share deemed too heavy. Handling hybrid war risk belongs to a wider cleanup, in which the market is trying to take back control of what it actually insures.

V.The Impasse of Attribution

To exclude is to name a culprit

The whole structure rests on a fragile premise, that when the time comes someone will be able to say an attack really came from a state. Yet attribution is the most difficult operation in the entire cyber field, and hybrid war is precisely designed to make it impossible. The clauses look first to attribution by the government of the state where the affected system is located. But governments rarely attribute attacks in public, because they do not want to expose their sources and they fear the diplomatic fallout¹⁰. Failing that, the clauses fall back on an objectively reasonable inference. That formula hands the insurer and the insured a call that states themselves avoid making.

This mechanism places the insurer in a singular, almost sovereign position. When no government attributes the attack, the insurer claims the right to decide that a state was behind it. It becomes the judge of a deeply political question, one that touches relations between nations and that foreign ministries handle with extreme care. To decline a claim, a London underwriter would have to settle what a foreign ministry will not say in public. Asking an insurance technician to make a geopolitical call weakens the clause. It also drags the insurer into endless disputes, since every attribution becomes something to litigate.

The gap between the loss and its attribution

Then comes the time lag, often ruinous. NotPetya was officially attributed to Russia only on 15 February 2018, eight months after the events, and the indictment of the responsible officers took three years. How do you settle a loss when its characterization, and so its coverage, hinges on an attribution that may come years later, or never? Deniability is the principal weapon of hybrid war, and it is also what disarms the exclusion. By refusing to declare itself, the adversary deprives the insurer of the fact on which the whole clause rested. In the Baltic cable cases, nobody can yet tell an act of war from plain maritime negligence. The impasse is playing out in real time.

VI.Toward a Public Backstop and the Limit of the Insurable

The precedent of terrorism

If hybrid war exceeds the market's capacity, the question becomes who will bear the burden. The history of insurance offers an illuminating precedent, that of terrorism. After waves of attacks that the market alone could no longer insure, several states set up public or mixed schemes of last resort, such as Pool Re in the United Kingdom, the TRIA program in the United States and GAREAT in France. Through them, the state takes on the catastrophic layer of risk that insurers cannot absorb¹¹. These mechanisms rest on a simple idea, certain risks are too correlated and too massive for private pooling, and only the community can be their insurer of last resort.

The question of the cyber backstop

The same debate has opened for catastrophic cyber. A federal backstop for the most devastating attacks has been floated as far up as the US National Cybersecurity Strategy, though it remains exploratory¹¹. The question picks up a recurring theme of this series, already seen with climate risk and catastrophe bonds. When a risk grows too large, the uninsurable share turns to public support. Seen this way, Lloyd's retreat from hybrid war draws a clear-eyed line where private insurability stops and collective solidarity has to take over.

Such a backstop would nonetheless run into the same obstacle as the exclusion it would complete. For a public scheme to be triggered, the event that activates it must be defined, and any definition of a catastrophic cyber war runs once again into attribution and characterization. Too broad, and the backstop subsidizes the least careful organizations, which creates moral hazard. Too narrow, and it leaves no relief for victims of an attack whose state origin could not be proven. Designing public support for cyber risk moves the core difficulty from the policy wording to the statute, and leaves it unresolved.

Insuring the knowable, not the ambiguous

The limit of what is insurable has always matched the limit of what is knowable. You can only price what you can measure, and only pay what you can characterize. Classical war, however terrible, remained knowable, and its exclusion workable. Hybrid war is entirely built to be unknowable, undeclared to escape the law of war, unattributable to escape responsibility, ambiguous to escape characterization. By withdrawing from this terrain, the market acknowledges that it cannot insure ambiguity itself, whatever the size of the risk. Better-drafted exclusions will not change much. Someone will have to be named to carry a risk whose author has made indeterminacy his strategy, and a society will have to decide how much of the war now hiding beneath the appearances of peace it leaves without a guarantor.

Corrected on October 31, 2026: NotPetya was attributed on 15 February 2018, eight months after the events, not nearly a year; the global cyber premium of about $9bn is the 2022 figure, not 2023; the B versions of the LMA clauses are not “set aside” by bulletin Y5381, they fall short of it absent Lloyd's prior agreement.