Insurance news

An AI agent that causes a data breach may trigger a business’s cyber policy. One that makes an unauthorised purchase or agrees terms with a customer could leave the same business with a loss that falls between covers.

OpenAI’s decision to cancel the October release of GPT-6.1 Astra has sharpened questions about what happens when an agent acts beyond its instructions. The Wall Street Journal reported that the model took actions to complete tasks without asking for permission during internal testing, and OpenAI has confirmed it will not proceed with the planned release. The Observer has also reported that insurers are assessing the risk of providing cover for damage caused by AI cyberattacks, with liability for rogue agents still unresolved.

George Grimshaw, divisional head of cyber and technology at Clear Group, said an agent acting outside its instructions does not automatically put the resulting incident beyond existing cover.

“Most cyber policies are triggered by what happens, not by who or what caused it,” he said. “If an AI agent goes beyond its instructions and that leads to a data breach, a system outage or a regulatory investigation, there’s a good chance a standard cyber policy responds.”

That depends on the wording not excluding the event, he said, and on the event fitting a defined trigger such as a security failure or system failure. It also depends on what affirmative AI language has been added to the policy. Grimshaw said insurers have moved to add that language across the market this year. He questioned whether naming AI exposures could introduce conditions that narrow cover which might otherwise have responded.

Where the loss falls between policies

The picture becomes trickier when an agent does something commercially damaging, such as making an unauthorised purchase, agreeing terms with a customer or sending the wrong information to thousands of people.

“Those losses can fall between cyber, crime and professional indemnity policies,” he said. “Crime cover typically needs a dishonest person or third party deception. Professional indemnity responds to claims from third parties rather than the business’s own losses. That gap is where businesses should be looking hardest at their programmes right now.”

Recovery depends on the developer’s contract

When a client suffers harm, Grimshaw said the business deploying the agent will usually be first in line. To the client, the agent is acting on that business’s behalf, and the business chose to deploy it. Whether it can then recover from the developer depends largely on the contract, and many AI providers cap their liability tightly or exclude it for how outputs are used.

Tim Johnson, partner and head of insurance and business and professional risk at Browne Jacobson, said courts have yet to provide clear answers on how liability will be divided between developers and the businesses deploying their agents.

“So it would depend on what loss has arisen and what’s actually caused it,” he said. “Is it a defect in the model itself, or is it a defect in the way it’s been used?”

Johnson suggested liability could rest with the developer where a model is used as designed, without modification. Where tools, prompts or other changes cause the loss, responsibility could instead fall on the user.

In some business-to-business arrangements, Johnson said, customers might accept an element of risk in return for a lower price.

“But equally, there are contractual mechanisms you may be able to deploy in some circumstances to mean that if the AI does go wrong, actually, you know, some of that risk may be laid off elsewhere.”

Proving what went wrong

Johnson said prompts and other relevant inputs would ordinarily be disclosed and scrutinised in a claim, helping courts assess whether the user had acted negligently or breached an obligation. How the model reached its output could be harder to establish.

“Was it the prompt that was wrong, or was it the AI itself going wrong?” he said. “And without having the ability to open the bonnet metaphorically [speaking] and see the thought process, I could see there being some really tough disputes over that.”

A cyber policy may respond to the breach an agent causes and leave an unauthorised purchase outside cover. If the developer’s contract also limits recovery, the business could retain that loss despite having insurance in place.

A former psychiatric hospital patient cannot pursue the company that later acquired the hospital’s operations for alleged abuse by staff, the Court of Appeal has ruled, in a decision with pointed implications for how liability and insurance track through business transfers.

The ruling, handed down on September 8, 2026, turned on a question that had never been squarely decided at appellate level in over four decades of TUPE law: does an employer’s vicarious liability to an outside claimant follow the employees to a new employer when a business changes hands?

The answer, the court held unanimously, is no.

What the claimant was up against

The claimant, identified only as ABC under a court anonymity order, alleged she was mentally and verbally abused and restrained more than 200 times during a four-month stay at Huntercombe Hospital in Maidenhead in 2018 and 2019. The privately run psychiatric facility was owned and operated by Huntercombe (No.12) Limited.

ABC initially sued Huntercombe alone. The problem was that Huntercombe had since gone into liquidation. It did carry public liability insurance – but with a deductible of £250,000 per claim. According to ABC’s legal team, that excess would likely consume most or all of any damages award.

So ABC changed tack. In around March 2021, Huntercombe’s operations had been transferred to Active Young People Limited under the Transfer of Undertakings (Protection of Employment) Regulations 2006 – the rules commonly known as TUPE. Two doctors who had at various times been ABC’s consultant psychiatrist and responsible clinician transferred with the business. ABC joined AYPL and the two doctors as additional defendants, arguing that Huntercombe’s vicarious liability for the doctors’ alleged acts and omissions had passed to AYPL through TUPE.

The stakes were not small. The court was told more than 50 similar claims by other claimants are pending against the same parties, with a case management conference scheduled for the autumn.

The legal question – and why it matters to insurers

TUPE exists to protect employees when their employer changes. Regulation 4(2)(a) says that on a relevant transfer, all the transferor’s rights, powers, duties and liabilities “under or in connection with” the transferred employees’ contracts of employment pass to the new employer.

ABC’s argument was straightforward: vicarious liability arises because an employee commits a wrong in the course of employment. That connection to the employment contract, ABC said, was enough to bring it within the sweep of TUPE.

AYPL countered that TUPE’s entire purpose was to safeguard the rights of employees – not to give outside claimants a new target when the original employer runs out of money.

Five reasons to say no

The Court of Appeal gave five reasons for dismissing the appeal.

First, the underlying EU Acquired Rights Directive – the source of the TUPE regime – exists to protect employee rights on transfer. Vicarious liability gives an employee no right they can enforce against their employer. It is, the court said, a secondary liability, always dependent on the employee’s own direct liability to the injured party. In the court’s words, it is simply a legal construct – a way for a claimant to pursue an employer likely to be better funded than the negligent employee.

Second, the so-called “protection” that vicarious liability offers employees is illusory. Under established authority, an employer who pays out on a vicarious liability claim is entitled to recover a full indemnity from the employee responsible. Far from shielding employees, the doctrine ultimately points the bill back at them.

Third, reading “in connection with” in the context of the Directive’s purpose, the connection between an employer’s vicarious liability to an outsider and the employee’s contract of employment is not the kind of connection the regulation contemplates. The driver for the relevant connection, the court held, must be the Directive’s purpose of protecting employee rights – and vicarious liability does nothing for that.

Fourth, the TUPE regulations read as a whole support that conclusion. Regulation 11 requires a transferor to give the incoming employer detailed information about employee-related liabilities before a transfer – claims brought by employees, disciplinary history, pending tribunal actions. There is no equivalent provision for third-party claims. If Parliament had intended those to transfer too, the court reasoned, it would have built in the same disclosure machinery. Its absence was not an oversight but a deliberate policy choice.

That point had real teeth on the facts. If AYPL were liable for 50-plus abuse claims worth potentially millions of pounds, it would have had no entitlement under the regulations to know they even existed before agreeing to the transfer.

Fifth, standing back from the technicalities, it struck the court as fundamentally counter-intuitive that a third party could bring claims against a transferee about events that happened before the transfer, about which the transferee knew nothing and had no right to know anything. ABC’s entire case, the court observed, depended on the happenstance of a business transfer that had nothing to do with her.

What the court said about the insurance position

Although it did not need to decide the point, the first-instance judge had indicated that if vicarious liability did transfer, the transferor’s right to claim on its public liability insurance would transfer with it. The Court of Appeal did not disturb that finding. Existing authority already establishes that an employer’s liability insurance covering liabilities connected with an employee’s contract – such as a personal injury claim by the employee – travels with the liability on transfer.

The practical upshot is that AYPL does not inherit Huntercombe’s liability, and ABC and the other 50-plus claimants are left pursuing an insolvent entity whose insurance deductible may swallow their claims.

What this means for the market

The court was careful not to criticise ABC or her advisors, acknowledging she may have suffered loss for which, through no fault of her own, she may not be able to recover. But it described the claim against AYPL as opportunistic.

For claims teams and underwriters working on business transfers, the ruling draws a bright line: TUPE moves employee rights and the liabilities that go with them, but an employer’s vicarious liability to outside claimants stays behind with the seller. Transferees cannot be ambushed by pre-transfer tort claims they had no way of knowing about – but equally, claimants chasing an insolvent transferor may find that the insurance sitting behind it is structured to offer little practical recovery.

A fraudulent umbrella company’s liquidators have failed to claw back more than £257,000 from two NHS consultant doctors who were paid without proper tax deductions – because the doctors had already declared everything to HMRC themselves.

The High Court dismissed both routes the liquidators tried in Re Fulmar Contracting Ltd (in liquidation) [2026] EWHC 2322 (Ch), handed down on September 9, 2026. The decision offers a detailed roadmap for how self-assessment by workers can defeat a transaction-at-undervalue claim – a question that matters directly to insolvency practitioners and the professional indemnity insurers behind them.

Fulmar Contracting Ltd was incorporated in March 2021 and wound up by September 2023, after HMRC presented a winding-up petition. The company had failed to pay very significant sums of PAYE, national insurance contributions, and VAT. It operated as an umbrella company, sitting between an employment agency and NHS workers. Its job was straightforward: employ the workers, deduct payroll taxes, pass them to HMRC, and pay the workers what was left. Instead, it paid the workers gross and pocketed the difference – or, more precisely, the people behind the fraud did.

The two respondents – both consultant doctors – were supplied with locum work through an agency called Fresh Medical. One was paid approximately £392,500 between September 2021 and May 2023, with part of that sum going to a company under her control rather than to her directly. The other received about £271,200 between August 2021 and July 2022. There is no evidence that the company deducted PAYE or national insurance from those payments, though one doctor’s pay was reduced for a period in a way that may have involved deductions of some kind.

The liquidators pursued two claims. The first was contractual: they relied on a clause in what they said were the doctors’ employment contracts, requiring repayment of any overpayments arising from the company’s failure to deduct tax. On their calculations, the two doctors had been overpaid by roughly £155,400 and £102,000 respectively.

The court found the employment contracts were never actually agreed to. The documents had been signed via DocuSign on the company’s side, but the doctors’ names were simply typed in – no ink signature, no DocuSign authentication mark for them. The doctors said they never saw the contracts, and the court accepted that evidence. Fresh Medical, the employment agency that provided copies to the liquidators, confirmed the contracts were obtained purely for its own internal compliance and were never sent to the doctors. The contracts themselves were riddled with errors: a key defined term – “Services” – was left blank, and the pay rate was set at the national minimum wage despite the doctors being paid at consultant rates.

The court also rejected the argument that the doctors became employees by conduct, finding no evidence of supervision, direction, or control – the common law hallmarks of an employment relationship – and nothing to suggest a clause like the repayment provision would have formed part of any informal arrangement.

The second claim was under section 238 of the Insolvency Act 1986, which allows liquidators to recover value lost through transactions at an undervalue. The argument was that by paying gross without deducting tax, the company gave the doctors more than they were entitled to receive, while simultaneously landing itself with a tax bill it could never pay. The extra amount the doctors received – the tax that should have been withheld – was the undervalue.

This is where the case broke new ground. The liquidators relied on Purkiss v Kennedy [2024] EWHC 1081 (Ch), where gross payments by an umbrella company to its employees were held at first instance to be transactions at an undervalue. The claims in that case ultimately failed on other grounds, and the Court of Appeal did not revisit the undervalue finding. But the court in Fulmar drew a sharper factual distinction: in Purkiss, the workers never declared the money to HMRC at all. Here, both doctors had filed self-assessment returns accounting for everything they received. One doctor’s declared self-employment income comfortably exceeded what she received from the company. The other’s returns, combined with the corporation tax filings of a company she controlled, covered the full amount.

The court held that by declaring the payments and making themselves liable for the tax, the doctors effectively eliminated the undervalue. HMRC had accepted the self-assessment returns without challenge. Allowing the liquidators to recover again would amount to double taxation of the same income – something the court said was inconsistent with both the self-assessment regime and the statutory purpose of section 238, which exists to reverse gratuitous depletions of an insolvent company’s assets.

The liquidators had offered a concession: they would give credit for tax the doctors had actually paid. But the court said the concession was too narrow. It should have extended to all income the doctors had declared to HMRC, not just the tax already handed over. The difference mattered because self-employed taxpayers can legitimately claim expenses and allowances that reduce the tax payable on the same gross income. The court refused to let the liquidators pick apart the doctors’ tax returns to challenge those deductions, finding the exercise incapable of proving anything relevant to the proceedings.

Even if it was wrong on all of that, the court added, it would have granted no remedy. The doctors were honest taxpayers who accounted for what they received. Had the company actually deducted tax properly, it almost certainly would not have passed the money on to HMRC anyway – that was the whole point of the fraud. HMRC’s position as a creditor was arguably better, not worse, because the doctors had done the right thing.

The judgment carries a careful caveat. Other workers paid by the company – or by similar umbrella companies – may still face valid claims. Each case turns on its own facts, and the court stressed that its decision does not resolve anything beyond the two claims before it.

For professional indemnity and D&O insurers, the decision maps out when a self-assessment defence can neutralise a transaction-at-undervalue claim – and when it cannot, where workers have not declared the income or HMRC has challenged the returns.

The parties were directed to agree a consequential order. The period for seeking permission to appeal runs from the date of that order.

Cyber underwriters trying to size up the next generation of AI may be about to lose one of their better sources for testing. The White House has asked OpenAI and Anthropic not to share their new models with the UK government’s AI Security Institute (AISI) until the US government has tested them, according to Politico, which cites a person familiar with the matter and a senior US administration official.

The request came from the Office of the National Cyber Director. The White House wants to make sure US systems are secure before the models are shared with partners. The White House, Anthropic and OpenAI did not immediately respond to requests for comment.

If the labs go along with it, AISI loses the privileged early access it has enjoyed until now, despite being one of the best-resourced government testing agencies in the world. That is a problem for the insurance market.

Independent pre-release testing is one of the few ways to see what the most capable systems can do before they reach clients’ networks. AISI has previously published joint pre-deployment evaluations with its US counterpart that tested models for cyber capabilities, biological capabilities, and software and AI development.

Access was already patchy. Anthropic recently released its latest model without giving it to AISI for testing beforehand. In a letter to MPs, AISI director Henry de Zoete acknowledged the gap. He said the institute still had “pre-release access to some of the world’s most capable models”, and noted that it tested OpenAI’s GPT-6 Astra before its public release. Anthropic has said it is working with the US government to open access to more domestic and international partners as quickly as possible.

Read next: Insurance is all in on AI, but the foundations are shaky

A break-in with nobody behind it

The request comes as the White House works out how to handle powerful new AI models that have hacked into outside organisations during testing. The latest case is in Australia.

Prime Minister Anthony Albanese said an OpenAI agent accessed non-public parts of a government Medicare statistics portal on 18 June. The agent kept going after being blocked; in Albanese’s words, it “didn’t accept no for an answer”. He said the agent also wrote data into the government’s database rather than only reading it, which raises the possibility that records were altered.

No personal information is believed to have been accessed, and a forensic investigation is under way. OpenAI did not tell Australian authorities until 10 September, nearly three months later. The company said the activity happened during an internal evaluation, while its models were trying to look up statistics about Australia.

For cyber wordings, the awkward part is that no human attacker was involved. Mark Luckin of Lockton Companies Australia told Insurance Business that better wordings apply an objective test of whether the insured authorised the access, rather than asking what the person or machine intended. “The trigger should be the outcome, not the motive behind it,” he said.

The UK has already seen the opposite problem. Ed Ventham of Assured Cyber told IB that cover struggled to respond to the UK Biobank incident because nobody had gained unauthorised access in the first place. The data had been accessed legitimately by approved researchers. IB’s analysis of both cases sets out the gaps they expose in current wordings.

Read next: AI agents are exposing the limits of the unauthorised access trigger

How carriers are responding

The market is moving, mostly by clarifying cover rather than excluding it. MSIG, QBE and Beazley are among the insurers reviewing traditional cyber policies to account for autonomous systems. QBE‘s global cyber head, Serene Davis, treats AI as an amplifier of existing risk: when an AI-related event causes a conventional incident, the loss still sits inside cyber cover.

Beazley says clients want AI folded into broad policies. Verisk‘s Jenny Soubra sees exclusions emerging at the edges in two areas. The first is systemic events, where one widely deployed model contributes to losses at many organisations at once. The second is cases where an agent working exactly as designed makes a costly decision on its own.

study covered by IB found insurers face hidden AI liability as agent risks multiply. It identified silent exposure concentrated in cyber, directors and officers, commercial general liability, and technology errors and omissions policies.

UK businesses are already feeling the effects. In QBE research on AI supply chain risk in UK cyber portfolios, 23% of UK businesses said they had suffered a cyber incident they believe involved AI. Cyber insurance take-up stayed broadly flat at 76%, against 77% in 2025.

Read next: Insurers face hidden AI liability as agent risks multiply

Westminster wants answers

MPs were already pressing the point before Washington stepped in. Liam Byrne, who chairs the Business, Innovation, Science and Trade Committee, has called senior people from OpenAI, Anthropic, Google DeepMind and Meta to appear on 13 October. MPs want to know whether safety testing should become mandatory and whether regulators should be able to block a model’s release. De Zoete will give evidence the same day. Byrne has said “voluntary self-regulation is unlikely to be a sustainable safeguard for the future”.

The UK’s financial regulators are working within existing rules for now. As IB reported when Silicon Valley’s AI extinction panic reached an already nervous market, the Bank of England, the FCA and HM Treasury warned in May that frontier AI models can now outperform a skilled human attacker on cyber tasks.

The labs themselves have called for governments to cooperate. OpenAI and Anthropic warned the UN Security Council about the risks of increasingly powerful AI and urged governments to work together on managing the technology.

Insurance Business has published its 2026 Brokers on MGAs report, identifying managing general agents (MGAs) that received high ratings from brokers across a range of service and performance measures.

The survey asked brokers nationwide to assess their MGA partners across 10 criteria, including niche-risk capability, pricing, responsiveness, technical expertise, and technology and automation.

The research provides a view of how brokers assess MGA performance across areas that influence placement, product access, service delivery, and relationships between intermediaries and MGAs.

How brokers rated MGAs

Insurance Business asked brokers to rate the performance and service of their MGA partners on a scale of 1 to 5, with 1 representing poor performance and 5 representing excellent performance.

The 10 criteria were:

• ability to place niche or emerging risks
• compensation, including commissions, bonuses, and profit sharing
• geographical reach
• marketing support
• overall responsiveness
• pricing
• range of products
• reputation
• technical expertise and product knowledge
• technology and automation

MGAs that achieved an average score of 4 or higher in at least one category received 5-star status under the report’s methodology.

The criteria cover both technical and commercial aspects of the MGA-broker relationship, including the ability to access specialist risks, product and pricing considerations, remuneration, and the use of technology in the distribution process.

Brokers’ Picks highlight product demand

The research also asked brokers to identify the top insurance products offered by MGAs.

The three products receiving the most votes were awarded the Brokers’ Pick medal, providing a separate measure of which MGA products attracted the most broker selections in the survey.

The report therefore considers both how brokers assess their MGA relationships and which insurance products they identify among the leading offerings.

Featured winners

Verve Risk Services and Geo Underwriting were among the MGAs featured as winners in the 2026 report.

The full report includes the list of winners, research findings, and wider industry insights from the broker survey.

The 2026 Brokers on MGAs report is supported by the Managing General Agents’ Association and is available to read for free, now.

The International Union of Marine Insurance (IUMI) used its annual conference in Rotterdam to debate whether freight forwarding liability insurance needs more systematic attention from the international marine insurance community.

Freight forwarders now play an increasingly central role coordinating global supply chains, but IUMI said this subline has received far less structured attention than cargo insurance.

“The freight forwarder is often at the centre of the supply chain organising and arranging all relevant activities for their clients globally,” said Matthias Kirchner (pictured), IUMI Executive Committee member and the workshop’s chair. “This is becoming increasingly a strategic role, but unlike cargo insurance, there is no globally consistent wording for freight forwarding liability insurance. Legal and regulatory frameworks vary considerably between countries, while insurers generally use their own individual policy wordings.”

Kirchner said IUMI could have a role addressing these challenges, potentially through a platform for sharing information.

A real gap, according to IUMI data

IUMI’s February 2026 webinar materials list global marine cargo premium at $22.6 billion for 2024, a well-established, tracked figure. The equivalent freight forwarding liability premium and loss data is marked simply “no data available.” That’s IUMI’s own presentation slide acknowledging the international market lacks reliable aggregate statistics for this line, exactly the problem Kirchner’s workshop discussed solving through an information-sharing platform and the collection of international market statistics.

Rotterdam builds on groundwork IUMI has already laid. The organisation ran webinars earlier in 2026 covering freight forwarders’ liability practices in China, Europe and the US, followed by a July session on India and France. Having mapped how differently the coverage works across major jurisdictions, IUMI is now weighing whether to push toward greater consistency.

Why a global standard is still far off

The workshop discussed whether a global freight forwarding liability insurance standard could eventually be developed, but IUMI said this would require significant further consideration given how much liability frameworks differ by jurisdiction.

Freight forwarders operate under a patchwork of national liability regimes, international conventions, and individually negotiated contractual terms, and insurers currently write coverage against their own bespoke wordings rather than any shared market standard.

The workshop flagged risk areas extending well beyond the physical movement of cargo: forwarders’ organisational structure, subcontractor oversight, documentation and information management, contractual arrangements, and incident and claims response.

The market is already moving on parts of this gap

Some of the innovation IUMI is discussing conceptually is already happening commercially. FIATA, the international freight forwarders’ federation, partnered earlier this year with Otonomi, an insurtech specialising in cargo delay insurance, to give FIATA’s member forwarders instant, algorithm-generated delay coverage quotes across air, ocean and e-commerce parcel shipments.

That product addresses business interruption risk tied to shipment delays specifically, and shows the market responding to freight forwarder-specific coverage gaps even without the standardised wording or aggregated statistics IUMI is discussing institutionally.

For insurers and brokers active in freight forwarding liability, this discussion signals the subline is likely to draw more institutional attention over the coming year, even if a formal standard remains distant given the jurisdictional complexity involved.

The more practical near-term outcome to watch for is a possible IUMI information-sharing platform, which would give underwriters a better basis for pricing and risk-managing a line that currently relies heavily on individual insurers’ own claims experience rather than shared market data.

London’s marine war-risk underwriters arrive at this week’s International Union of Marine Insurance (IUMI) annual conference in Rotterdam bringing with them a pretty big number.  Somewhere between $1.5bn and $2bn in claims from roughly 70 vessel casualties since the Gulf conflict began. One broking estimate suggests the final bill could yet exceed an entire year’s global premium income for the class.

That figure comes from IUMI secretary general Lars Lange, ahead of a conference that was opened by IUMI president Frédéric Denèfle. The conflict began with US and Israeli strikes on Iran on February 28 and is now closing in on seven months old.

For the Lloyd’s Market Association (LMA), which represents the underwriters pricing this risk, the position hasn’t shifted much since it started: the market has stayed open throughout, but the price reflects a war with no clear end.

“Cover available for a price”

Neil Roberts, the LMA’s head of marine and aviation, has repeated the same line whenever asked why ships were avoiding the Strait of Hormuz: there was always sufficient capacity, with cover available for a price. The LMA has resisted the suggestion, made at points by government and shipping-industry figures, that unaffordable or unavailable insurance rather than crew safety was keeping vessels in port.

The price itself has moved a long way. Hull war-risk rates for Hormuz transits started the conflict at around 0.15%–0.25% of vessel value. They spiked as high as 10% during the worst fighting and now sit at roughly 5% for the most exposed voyages. On a $150m tanker, that’s a $7.5m bill for a single transit.

Read next: Hormuz war-risk rates surge again as ceasefire collapses

Rate rises this steep tend to invite a familiar accusation, that insurers are profiting from the crisis. Lange’s figures, echoed by Lloyd’s own half-year numbers, tell a totally different story. Lloyd’s disclosed the scale of the conflict’s impact for the first time in its half-year results: £1.4bn, or about $1.9bn, in losses tied to the Middle East conflict, concentrated in marine, energy and political violence lines. Chief executive Patrick Tiernan said the exposures seen so far didn’t look like a capital event for the market. Pre-tax profit still fell 16.8% to £3.5bn for the half.

Broking analysis from Howden Re goes further. It puts total potential war, terror and political violence claims from the conflict at $2bn–$3bn against an estimated annual global premium pool for the class of just $1.5bn–$2bn. A single conflict, on that estimate, could cost the market more than a full year’s income from the business it’s meant to fund.

Read next: Lloyd’s puts a number on the Iran conflict for the first time – and flags a tougher year ahead

A soft market despite the losses

What makes the Gulf numbers harder to square is that they’re landing on a wider marine market that, by IUMI’s own account, isn’t hardening. Opening the Rotterdam conference, Denèfle described overall conditions as broadly stable but still predominantly soft, with much of the hull and cargo premium growth reported this year coming from currency movements – a weaker US dollar inflating figures reported in dollar terms rather than genuine rate increases. Offshore energy, he said, remains subdued.

His explanation for that mismatch is that war risk is rising, but so is competition from new capacity entering the market, alongside persistent inflationary pressure and continued uncertainty over trade and tariffs. Those forces are pulling in opposite directions at the same time, which is part of why a conflict big enough to eat a year’s premium income for one class hasn’t been enough to turn the broader market. Denèfle’s own summary of where that leaves insurers was that our sector remains strong, agile and ready to adapt.

That backdrop matters for brokers placing business outside the war-risk niche as it suggests renewal conversations on the standard hull and cargo book are unlikely to toughen just because Gulf headlines are worsening.

The real exposure is still to come

Even the $2bn figure under discussion in Rotterdam probably understates where this ends up. Many marine war policies now carry a 12-month waiting period before “blocking and trapping” cover responds for vessels stuck in the Gulf, treating a ship as a total loss if it stays stranded for a year, even undamaged. Hostilities resumed in August after a ceasefire collapsed within weeks, and vessels remain stranded on both sides of the Strait. A meaningful share of this war’s cost may not surface until well into 2027.

The Strait carries roughly a quarter of the world’s seaborne oil trade and around a fifth of global LNG shipments, according to the International Energy Agency. A prolonged closure reaches well beyond marine underwriters into energy pricing and the wider political violence and terrorism market.

Denèfle flagged a related, longer-term shift which is that insurers will increasingly be asked to cover longer trade routes that exist specifically to avoid conflict zones, as shippers reroute around the Gulf rather than pay war-risk premiums to cross it. That’s a different underwriting problem to the one dominating this week’s headlines, and one likely to outlast the current conflict.

Read next: Hormuz war-risk rates face fresh pressure as Iran-US clashes resume

Washington’s own fix still hasn’t written a policy

Throughout the conflict, one thing hasn’t changed: the US government’s attempt to solve it with its own insurance scheme hasn’t got off the ground. President Trump ordered the US International Development Finance Corporation (DFC) on 3 March to set up political risk cover for Gulf shipping. Within days the agency unveiled a facility worth up to $40bn, naming Chubb as lead underwriting partner. Six months later, reporting has repeatedly found the scheme has placed zero dollars of actual coverage. Senator Jeanne Shaheen has pressed the DFC on how taxpayer money is protected and who ultimately benefits from a facility built to keep the Strait open.

That leaves the commercial market carrying this risk on its own, at prices the LMA maintains reflect genuine danger rather than any shortage of capacity even as the wider market Denèfle described stays soft around it.

Read next: US senator presses DFC on taxpayer risk in $20 billion maritime reinsurance proposal