RESEARCH DOSSIER
Cyber Insurance Other Insurance and Contribution Priority
Abstract. Cyber insurance is frequently purchased within an existing enterprise insurance program that also contains commercial crime, property, general liability, professional indemnity, and technology errors and omissions coverage. The resulting interaction raises questions about simultaneous applicability, contractual priority, the insured's recovery, and the distribution of indemnity and defense costs between insurers. This monograph examines the legal and economic organization of those questions through comparative analysis of selected United States, English, Canadian, Australian, and German authorities, actual cyber policy wording, supervisory materials, and the cyber-risk literature. It distinguishes the factual identity of an incident from the legal identity of an insured loss and examines the significance of policy triggers, exclusions, the insured interest, competing other-insurance provisions, genuine excess layers, retentions, sublimits, and recoverability. Reproducible hypothetical allocations illustrate the consequences of equal-share, limit-proportional, and priority-based approaches. The synthesis establishes that the availability of insurance to an insured and the ultimate allocation of expenditure among insurers are separable questions, and that the outcome depends on the legal and contractual circumstances of each loss component. The evidence supports neither a universal cyber-first rule nor a single international contribution formula. Aggregate market statistics establish context, while the prevalence and causal effects of contribution disputes remain empirical questions requiring linked claim-level records. This study supplies an interdisciplinary examination of coverage coordination rather than a jurisdiction-independent prescription for claim settlement.
Keywords: Cyber insurance; other insurance; double insurance; equitable contribution; coverage priority; commercial crime; property insurance; professional liability; defense costs; contractual allocation; claims management; silent cyber; insurance regulation; risk accumulation.
Introduction
A malicious email, compromised identity, unauthorized transfer, interrupted service, or disclosure of personal information can produce several legally distinct losses. Some losses are costs incurred in a firm's own operations; others are liabilities owed to persons whose property, services, or information have been affected. An insurance program assembled over time may respond to particular portions of both. The existence of a cyber event is consequently a starting point for investigating insurance, not an answer to the allocation question.
The subject of this monograph is the coordination of direct insurance contracts where at least one is cyber insurance or provides affirmative cyber coverage. Other-insurance provisions address the relationship between a contract and another insurance obligation. Contribution concerns the redistribution of a shared insurance burden after, or in connection with, performance. Priority concerns when an obligation must respond relative to another. These questions interact but require different evidence. The same incident can trigger two policies covering separate loss components without generating double insurance; it can also give rise to a single indemnifiable liability that both policies cover.
The research asks how concurrent coverage arises, how competing clauses are interpreted, what distinguishes contribution from other recovery mechanisms, and what financial consequences follow from alternative allocation decisions. The scope is international and comparative rather than exhaustive. The analysis selects identifiable authorities in England and Wales, several United States jurisdictions, Canada, Australia, and Germany. Dates and jurisdictions attach to each legal proposition. The text employs American English, retaining jurisdiction-specific legal terms where material. The research cutoff is November 18, 2026; it does not represent a comprehensive citator or all-jurisdiction legislative update.
The methods are comparative doctrinal reading, structured analysis of a specimen policy, consideration of official supervisory and market materials, review of accessible research abstracts, and deterministic numerical illustrations. The cases are analytically selected because they expose different mechanisms; they are not a probability sample. The full judgment is used where accessible, and a published case summary or extract is expressly identified where full reasons are unavailable. Synthetic claims are constructed exclusively to expose the arithmetic of specified assumptions. Monetary amounts are hypothetical except where attributed to actual judicial or market materials.
The underlying scientific context is cyber insurance as a financial response to losses arising through information systems, technology dependencies and digital interaction. Biener, Eling and Wirfs (2015), in the accessible abstract of their empirical study of insurability, identify dependence, information problems and limited data as obstacles to sustainable transfer of cyber risk. Eling and Schnell's (2016) review of cyber risk research locates insurance questions within a wider research field. This study extends the investigation to the organization of multiple simultaneously relevant contracts, without claiming discovery of a new general theory.
Conceptual and Historical Foundations
The unit of analysis
The unit of loss must be more precise than the event label. Suppose an employee acts on a false payment instruction, the organization loses funds, its email service is compromised, and customers discover that their confidential information was exposed. Each outcome has its own factual origin and valuation. An invoice for forensic services, a lost bank transfer, a contractual damages award and a settlement of privacy litigation are distinguishable expenditures. Some may be compensable by more than one policy; others may fall outside every relevant agreement.
An insured interest describes the relationship between an insured and the subject matter of a possible loss. A company may own funds, hold customer funds in trust, operate servers, owe a contractual performance obligation, or face liability arising from a security failure. Insurance of one interest does not automatically cover another. The property of a managed service provider does not become the customer's property merely because both suffer consequences of a shared breach. Likewise a parent and subsidiary can be insured under the same program while having different losses and policy entitlements.
For the research purpose, a candidate concurrent obligation can be described by identifying the legal insured, the insured interest, the loss component, the causal facts, the relevant dates, and the particular contract. These items form an evidence record, not a proposed legal test applicable in all countries. The legal obligation is established under the actual wording and governing law. The factual variables help ensure that the analyst applies the appropriate language to the appropriate loss.
A security event may be measured by its technical footprint, duration and origin. An insurance claim may be measured by the financial loss alleged, the amount reserved and the amount paid. A contribution action may be measured by the amount advanced by one insurer beyond a later determined share. These measures have different denominators and cannot be interchanged. In particular, one underlying incident may result in multiple claims, including several claims within the same policy and multiple policies held by the same organization.
The historical path to overlap
Insurance contracts were organized around classes of loss long before the emergence of the contemporary standalone cyber policy. A network-based loss can therefore enter established insurance through provisions concerned with theft, damage, interruption, or legal responsibility. Standalone cyber insurance adds coverage expressly designed around information-security and privacy-related events, often together with related response services. As a result, old and new contracts coexist in the same program.
In its 2021 Cyber Insurance Risk Framework, New York's Department of Financial Services identifies non-affirmative exposure in several existing insurance lines. The institutional issue is the uncertainty of exposure when wording does not make the treatment of cyber-related loss explicit. The regulator also addresses the concentration of technology dependencies. These concerns illuminate the origins of coverage overlap but do not establish that an insurer must cover every cyber-related loss under a traditional policy.
The replacement of ambiguous wording with express grants or exclusions can clarify what a contract covers. Even a market consisting entirely of affirmative grants may retain concurrent coverage. For instance, two contracts can deliberately insure electronic theft while treating different maximum amounts or conditions. Conversely, two policies can each expressly address cyber risks yet leave a gap because their respective triggers do not coincide. Affirmation and exclusion concern contract scope; coordination concerns the relationship between actual obligations once scope is determined.
The historical expansion of specialized coverage has thus added an additional dimension to insurance program design. A broad commercial program can contain simultaneous first-party and third-party coverage, different time triggers, shared or separate aggregates and insurance provided to additional insureds. The scientific inquiry must follow these arrangements through specific contract obligations rather than assign a fixed hierarchy to product names.
Horizontal and vertical risk transfer
Horizontal coverage refers here to policies capable of responding at the same attachment level to a loss, whether intentionally purchased as concurrent primary coverage or treated as such after applying relevant legal rules. Vertical coverage attaches at a specified level above underlying insurance. A specialty contract may include several coverage sections with separate coordination rules. The description of a whole policy as primary does not establish the status of every coverage section.
A true excess contract is purposefully designed to respond above an underlying layer; its schedule, attachment provisions, limits and exhaustion conditions determine the obligation. A primary policy with a contingent excess clause can perform as primary when there is no qualifying other insurance. It is analytically dangerous to identify the two arrangements merely because they both use the word excess. One is part of the intended layer architecture; the other can become relevant only when a condition of overlap is satisfied.
Coinsurance within one placement, in which insurers subscribe percentages of a common layer, differs from separate overlapping insurance contracts. A subscription agreement ordinarily supplies a stated allocation from the outset. Independent policies require investigation of whether an obligation is common and how it is allocated. Reinsurance, in turn, addresses the insurer's own transferred risk and normally operates through a separate contractual relationship. Its presence does not by itself create a second direct policy for the original insured.
Indemnity and the problem of multiple rights
An indemnity contract commonly responds to a defined compensable loss, subject to its terms. Where the same indemnifiable loss is covered more than once, insurance may expand recovery capacity without permitting recovery beyond the loss. A firm with two valid policies may have more routes to obtaining payment, greater available limits, or protection against an exhausted policy. Those benefits have economic value even when the final recovery is limited to the loss.
This distinction is central to the analysis. Double insurance is not synonymous with double recovery. Multiple independently purchased policies may supply the resources for a single full indemnity. An insurer that pays a covered loss can have a separate right of contribution from another insurer, depending on governing law and the relationship of their obligations. The procedure that protects the insured from underpayment can coexist with a procedure that distributes the ultimate burden fairly among insurers.
The actual indemnity amount is not always equal to gross economic loss. Deductibles, coverage exclusions, valuation conditions, waiting periods and limits modify the compensable amount. Furthermore, defense costs may be covered on a different basis from damages. A useful financial ledger therefore preserves gross loss, covered amount, unpaid retention, covered defense expenditure and remaining available limits as distinct quantities. An apparent surplus disappears when these distinctions are applied to the amounts in question.
Primary Documents and Scope of Other-Insurance Clauses
Coverage before coordination
The first interpretive question is whether each insurance agreement independently applies to the circumstances. Relevant features include the definitions of insured and claim, applicable period, retroactive date, property or funds covered, causation language, insuring agreement, exclusions and compliance with conditions. A provision that determines priority cannot substitute for an affirmative grant of coverage.
For each purportedly overlapping obligation the analyst needs the declarations, form, all endorsements, changes of insurer or insured and any negotiated manuscript endorsements. A publicly accessible specimen establishes possible language for analysis but cannot establish the wording governing a particular claimant. The order of precedence among form, endorsements and schedule can itself determine the outcome. Distinct editions should be preserved as separate historical evidence.
The temporal dimension is equally significant. An event can occur in one year, become apparent in another, generate a demand in a third, and ultimately result in payment years later. An occurrence policy, a discovery-based crime policy and a claims-made policy may refer to different milestones. Overlap cannot be assumed solely from calendar overlap of stated policy periods. One must establish the applicable trigger for each contract. A contractual relation-back or related-claims clause may associate later demands with earlier notices; another policy may use a different aggregation unit.
An actual standalone cyber form
The At-Bay (2023) cyber policy form AB-CYB-001.2, edition August 2023, supplies concrete language for examining coordination. Section VII.J.1 makes the policy excess over other qualifying insurance for covered amounts and carves out insurance expressly written above its own aggregate or sublimits. The interaction matters: an apparently broad excess rule is accompanied by a qualification designed to preserve a deliberately arranged vertical layer. The text therefore demonstrates why the other-insurance condition must be interpreted with the actual excess placement.
Sections VII.K and VII.L separately address recovery rights following payments and the order in which external recoveries are applied. These provisions concern different financial relationships from the other-insurance clause, even though all can affect the insurer's net cost. The specimen further illustrates the need to inspect reporting provisions together with coordination language, since the availability of later purchased insurance can bear on certain discovery-reporting or notice-of-circumstances conditions.
The form is one policy edition. Its legal effectiveness, the interpretation of collectible insurance, and the validity of its application in any individual case depend on the applicable insurance contract and local law. It cannot be treated as a statistical description of the cyber insurance market. A meaningful policy-form survey would sample versions by jurisdiction, year, distribution channel, insured size and coverage section, recording whether endorsement language changes the base wording.
Clause functions
A pro rata clause states a method or principle of shared allocation when other qualifying insurance also applies. The operative denominator may be policy limits, amounts of liability, shares specified in the contract, or another measure. An equal-share provision may specify successive contributions until smaller available amounts have been exhausted. The mathematical distinction becomes important when policy capacities are unequal.
An excess other-insurance clause seeks to postpone payment relative to qualifying other coverage. Its effect depends on the exact phrasing, which insurance counts as other insurance and the law governing competing clauses. A policy may expressly limit its treatment of other insurance to the same loss, an explicitly listed risk, or a specific underlying agreement. Some wordings address deductibles under other coverage or the consequences of uncollectibility; these details alter both the priority and the available amount.
An escape clause seeks to eliminate or substantially restrict liability where other insurance is available. A noncontribution provision can be framed in different ways, including excess positioning. Labels are convenient descriptions of possible functions rather than self-executing legal classifications. The actual language determines whether clauses conflict, can be reconciled, or are ineffective under an applicable statute.
A primary-and-noncontributory endorsement may implement requirements arising from a commercial relationship, for example where a customer is included as an additional insured under a vendor's policy. Its operation depends on the scope of additional-insured coverage, the identity and capacity of the parties, and the wording of competing insurance. A contractual indemnity between commercial parties may influence their own liability or recovery relationship; it does not rewrite a different insurer's contract without a legal or contractual basis.
Overlap between cyber and commercial crime
Business email compromise demonstrates how the same transfer may be analyzed under several insurance grants. A crime policy may cover computer fraud or another specified form of theft, while a cyber policy may include a fraudulent instruction benefit. Ownership and control of funds, the mechanism of the transfer, the identity of the instructed person and the sequence of approvals can determine whether the respective definitions are satisfied.
In American Tooling Center v. Travelers (2018), the Sixth Circuit, applying Michigan law, reversed the district court's judgment for the insurer in a computer-fraud dispute. The decision concerns the particular wording and causal circumstances. It demonstrates how a digitally facilitated monetary loss can be treated under a non-standalone-cyber contract; it is not a determination of priority between two cyber-related policies.
A crime policy may cover a transfer without covering technical remediation. A cyber policy may cover security investigation without covering the stolen funds. A payment to restore system access and a payment sent to a fraudster pursuant to a deceptive invoice also involve different factual and legal questions. Where two policies cover the transferred funds but only one covers forensic expense, contribution analysis must distinguish those categories.
The financial investigation starts with bank records, transaction authorizations, communications, ownership evidence and amounts recovered. A returned transfer reduces the economic loss; the treatment of recovery expenses and retention depends on the contract or governing law. An invoice that represents costs attributable to incident response should not also be entered as a separate item of stolen property.
Overlap with property and interruption coverage
An incident involving corrupted information, inaccessible systems or interrupted production can be investigated under traditional property coverage as well as affirmative cyber coverage. The specific trigger, definition of covered property, valuation and any explicit cyber exclusions are essential. A cyber interruption benefit may operate without the same property-damage prerequisite as a property-based business interruption contract.
The New Jersey Appellate Division considered a legacy property-program war exclusion in Merck v. ACE (2023), in litigation arising out of NotPetya. The court upheld the relevant lower-court determination concerning that exclusion. Its holding concerns the policy wording before the court. The litigation is useful here as evidence of the potential connection between technology-related losses and established property insurance, not as a precedent allocating a payment between a standalone cyber insurer and a property insurer.
If a property policy responds to damage of a particular insured asset and a cyber policy responds to the cost of restoring data on it, the analyst must establish whether those expenditures actually duplicate each other. A replacement device and an invoice for recovering information may be distinct. If both policies value exactly the same replacement cost, the possible concurrent obligation is different. When a business interruption calculation includes expenses saved by shutting operations down, those savings affect the net loss; the physical interruption and economic loss must not be counted twice under separate policies.
Overlap with liability and professional insurance
An information-security incident may produce privacy claims, contractual allegations, professional negligence allegations and claims for failure to deliver a technology service. A standalone cyber liability grant and a professional indemnity grant may cover the same civil liability or different duties owed by the insured. General liability may present another possible grant, depending on wording and exclusions. In claims-made programs, notice and aggregation interact with these grants.
The distinction between damages and defense expenditure is material. One policy may provide reimbursement of defense expense, while another imposes a broader obligation to defend qualifying claims. Defense may be inside or outside the limit; counsel selection, consent, settlement cooperation and allocation for uncovered causes of action may vary. The fact that one insurer funds counsel does not alone prove another insurer bears the same defense obligation.
Each claimant, alleged wrongful act, cause of action and category of expenditure must be associated with the relevant policy technology. An aggregation provision can consolidate multiple related demands for purposes of one limit even when another insurer counts them separately. The coexistence of aggregation provisions can therefore change the apparent size of the overlapping layer.
Mechanisms and Evidence for Contribution
The separate claims
The policyholder seeks performance under contracts of insurance. A participating insurer may seek contribution for a payment discharging a common insurance obligation. An insurer may also seek subrogated recovery from a legally responsible third party, such as a tortfeasor or contractual indemnitor. These rights can coexist but their predicates, defenses, limitation periods and distributions need separate analysis.
In Fireman's Fund v. Maryland Casualty (1998), the California Court of Appeal recognized equitable contribution as an insurer's independent right where applicable primary obligations cover the same insured and risk. The ruling also examined allocation over overlapping periods on the particular record. It rejects the necessity of recharacterizing that independent claim as the insured's subrogated right. The decision should not be read as resolving all priority disputes between primary and true excess insurance.
The Maryland appellate decision in Selective Way v. Fireman's Fund (2023) likewise analyzes contribution in relation to common defense costs and actual discharge of a shared burden. The reasons illustrate that the contribution claimant must substantiate its entitlement and that the procedural history of other proceedings matters. Neither case establishes a universal ratio for cyber loss. Both offer pertinent distinctions for interpreting common obligations in a new technological context.
A subrogated action generally relies on a right the insured possesses against another responsible person. For example, a vendor's liability for failure to perform agreed security services may support a separate recovery after insurance payment. The vendor need not be a concurrent insurer of the customer's loss. Conversely, an insurer's obligation to contribute can exist even though it did not cause the original security failure. This separation prevents a claim about sharing insurance liabilities from being confused with a claim about causation of the insured loss.
Competing excess conditions
Assume two independently purchased policies each contains an otherwise applicable primary grant, but each seeks to defer payment if another valid policy responds. If each clause is applied without regard to the other, a circular priority result can emerge: both insurers assert that the other should satisfy the first layer. The legal system must determine whether one clause has priority, both are neutralized, or another statutory or contractual rule resolves the dispute.
The Canadian litigation Family Insurance v. Lombard (2002 SCC 48) concerns incompatible excess conditions in overlapping liability contracts. The reported decision restores equal sharing in the circumstances before the court. The official reasons are identified, although the complete opinion was not available for detailed review in this research. Accordingly, the discussion uses the reported holding as a comparative illustration rather than an exhaustive account of Canadian conflicts doctrine.
The contrast with a genuinely purchased excess layer is important. If the second insurer contracted only to respond above a specified underlying amount, the two obligations do not necessarily exist at the same level. Treating every excess reference as a competing primary clause could undermine the intended underwriting arrangement. The schedule of underlying insurance and the requirements for exhaustion constitute distinct evidence.
Circularity can also arise through different wording. One contract may purport to be excess over all other insurance regardless of how that insurance is described; another may purport to be primary except for expressly specified insurance. A clause comparison should therefore record both the terms used and which obligations they cover. Even a superficially symmetrical conflict can have an asymmetrical resolution if one clause has different scope or a statute operates on only one.
The English cyber-specific dispute
Watford Community Housing Trust v. Gallagher (2025) is especially relevant because it explicitly addresses overlapping cyber, commercial combined, and professional indemnity arrangements following a data breach. The accessible introductory reasons and policy extracts describe a preliminary issue in a professional negligence proceeding against the broker. The first two reproduced contracts contained competing provisions subordinating coverage to qualifying alternative insurance while recognizing specifically excess policies.
The full judgment's concluding reasons could not be retrieved directly in this session. The published case summary reports that the court neutralized the competing conditions and distinguished the insured's aggregate recovery capacity from insurers' rateable contribution. Simmons & Simmons' commentary corroborates the reported horizontal layer of £1 million, £5 million, and £5 million, with additional defense-cost cover identified under the professional indemnity policy. The £11 million figure is conditional aggregate capacity in the counterfactual addressed by the court; it is not evidence that this amount was ultimately paid or that all claims were quantified at that amount.
The importance of the case for cyber insurance arises from the combination of policy categories and the broker-notification issue. It demonstrates the practical significance of preserving claims under every potentially responsive contract. It also demonstrates why a highest-limit ceiling and a prohibition of overindemnification are distinct concepts. Multiple limits can supply a larger recovery capacity than any individual limit without allowing recovery beyond the compensable loss.
This discussion confines itself to the reported preliminary determination. It does not infer a final broker damages award, the treatment of every policy term, or subsequent appellate history. Any claim relying on precise language in the unavailable part of the judgment requires direct verification of that passage. The accessible extracts and two summaries provide a bounded account adequate for identifying the mechanism, but not for asserting a comprehensive statement of English law.
Statutory treatment in Australia
Australia provides an important comparative alternative because the legislature has expressly addressed other-insurance provisions. The Insurance Contracts Act 1984 (compilation October 8, 2026), section 45, invalidates a provision to the specified extent that it limits or excludes liability because the insured entered other insurance, subject to the statutory qualifications for legally required insurance and specified underlying cover. This places the statutory test before any assumption that the commercial excess condition is enforceable.
Section 76 addresses rights against multiple insurers liable under separate general insurance contracts for the same insured and loss. Subject to its limits, the insured can recover from one or more liable insurers up to indemnity, and rights of contribution between insurers are preserved. The provisions therefore make the separation between recovery and allocation explicit.
The High Court of Australia's official description of Zurich v. Metals & Minerals (2009) identifies the significance of whether the insured entered the second contract or instead was a nonparty beneficiary. It also identifies the question of the extent to which a provision is rendered void. The source reviewed is the court's case description, not its complete reasons. A detailed application to a particular additional-insured arrangement would require examination of those reasons and the exact statutory and contractual conditions.
This qualification is significant for enterprise cyber insurance, in which technology contracts can confer insurance benefits on customers or other group entities without making them contracting policyholders under every agreement. The insured's role in procurement cannot be inferred from the fact of insurance protection alone. The scope of the relevant statute and contractual relationships therefore has to be mapped before evaluating priority.
Statutory treatment in Germany
Section 78 of Germany's Insurance Contract Act sets out a specific multiple-insurance rule. The official English translation identifies the same interest insured against the same risk, aggregate insurance exceeding the relevant value or loss, liability of each insurer in accordance with its contract subject to the aggregate loss ceiling, and inter-insurer allocation in proportion to the amounts for which the respective contracts make them liable. It also addresses the condition for compensation claims when foreign law governs one policy.
This provision offers a useful comparison with the common law examples. Liability amounts rather than nominal headline policy limits are the relevant basis for the statutory inter-insurer proportion. Differences between contractual payment obligations can arise through a sublimit, a deductible, a covered valuation or other terms. One cannot replace this statutory quantity with a limit ratio solely for computational convenience.
The English statutory translation is used for accessibility. The German statutory text, case law, conflict rules and effective legal version must control a real German dispute. The analysis does not treat section 78 as an international choice-of-law rule or assume that every insurance contract in a cross-border program is governed by German law.
Defense allocation and payment sequence
Indemnity allocation determines who ultimately bears the covered settlement or judgment. Defense allocation concerns the costs of responding to the claim. The same policy can provide different obligations for these expenditures. A defense outside the limit may leave the damages capacity intact; defense within the limit can reduce the remaining capacity substantially. If a policy is excess and has no current defense obligation, it may not owe the same defense costs as a primary carrier.
Payment sequence is the chronology of cash transfers. Priority is the contractual or legal ordering of obligations. An insurer may advance covered expenditure while other insurance disputes remain unresolved. When an applicable contribution right permits subsequent adjustment, the initial outlay can differ from its net ultimate expense. This is a consequential distinction for the policyholder's liquidity and the insurer's reserving.
The determination must remain specific to the operative policy grants and local law. Arrangements for interim cost sharing can specify reservation of unresolved rights and accounting procedures. Such arrangements may reduce delay without establishing how a court would allocate the eventual claim.
Comparative Measurement of Allocation Disputes
Measurement structure
The following deterministic calculations are original illustrations using hypothetical amounts. They are not calibrated actuarial models, forecasts, or universal legal rules. The assumed allocation method is an input, not an empirical finding. All examples use dollars in thousands except where stated otherwise. All policies are assumed to be solvent and enforceable unless another assumption is explicitly introduced.
Let D represent the single compensable loss to be allocated after applying the assumed common loss valuation, and let C_i denote the maximum payment available from policy i for that loss before application of the chosen allocation method. For a set of independently applicable indemnity obligations, any feasible payment vector x must satisfy 0 <= x_i <= C_i, and the total of x cannot exceed D. If all capacity can be accessed under the assumed legal arrangement, the total payment is the smaller of D and the sum of available capacities.
The word capacity refers to the amount available under the assumptions, not a policy's original declared aggregate. Deductible mechanics, waiting periods, exhaustion, defense erosion and contractual exclusions are ordinarily calculated before identifying it. Some contracts make capacity dependent on how much another policy pays, in which case these simple independent capacities cannot be assumed.
Equal shares and limit-proportional shares
Consider two policies with applicable capacities of 200 and 600 and a common compensable loss of 300. With equal sharing each pays 150. With proportional sharing using those capacities as the agreed weights, the first pays 75 and the second 225. Both distributions indemnify the same loss fully. Their differences concern who bears the financial burden.
| Assumption | Policy A payment | Policy B payment | Total |
|---|---|---|---|
| Equal shares | 150 | 150 | 300 |
| Proportional to available capacity | 75 | 225 | 300 |
| A primary, B excess above 200 | 200 | 100 | 300 |
| B primary, A excess above 600 | 0 | 300 | 300 |
The final two rows assume genuine operative priority, not a judicial inference from the product names. Equal sharing can be expressed by selecting a common payment level t such that the sum of min(t, C_i) equals the target payment. In the example, t is 150. If D increases to 500, both initially share 200 and policy B supplies the remaining 100, resulting in 200 and 300. If D increases to 900, both exhaust their capacities and the unpaid portion is 100. This piecewise calculation makes explicit how equal sharing changes when a smaller limit binds.
Capacity-proportional sharing under independent capacities is calculated as x_i = T C_i / sum(C_i), where T is the smaller of D and total capacity, provided the denominator is positive. With capacities of 200 and 600, the weights are one-quarter and three-quarters. At D = 500, the payments are 125 and 375. At D = 900 they are 200 and 600. The formula is one possible computation under its stated assumptions, not a claim about which formula governs German, Canadian, English or American contracts.
When the basis is the amount each policy would separately owe on the same loss, C_i should be defined accordingly, and nominal declared limits may be inappropriate weights. A clause might instead fix weights by limits even when other restrictions reduce feasible payment. Additional rules would then be required to redistribute a capped share or explain any remaining uninsured amount. The arithmetic follows the contract; the contract does not follow the arithmetic.
Disjoint and partially overlapping losses
Consider a hypothetical incident with the following loss ledger: transferred funds 420; forensic expenses 80; operating interruption 200; third-party damages 300. The total is 1,000. Suppose crime and cyber each initially cover the transferred funds subject to capacities 300 and 150. Only cyber covers the forensic expenses; the interruption is covered by property insurance; the liability is covered by professional insurance. Assume sufficient distinct limits for the last three payments and no retentions or exclusions.
Under proportional sharing for the funds, crime bears 280 and cyber 140. Cyber also bears 80 of forensic expenses. Property bears 200, and professional insurance bears 300. The total is 1,000; the insurer payments are 280, 220, 200 and 300 respectively. Equal sharing for the funds with the cyber capacity capped at 150 would instead allocate 270 and 150, resulting in 270, 230, 200 and 300 for the four insurers.
| Loss component | Gross amount | Candidate insurance in illustration |
|---|---|---|
| Transferred funds | 420 | Crime and cyber |
| Forensic response | 80 | Cyber |
| Business interruption | 200 | Property |
| Third-party damages | 300 | Professional liability |
A ratio calculated over the entire incident would wrongly distribute expenses that do not share an insurance obligation. Applying one aggregate limit across all cyber grants instead of the separate assumptions above can change feasibility. For example, if cyber's total remaining aggregate is 180, its 80 of forensic payment leaves 100 for transferred funds. With crime limited to 300, the funds recovery can be no more than 400, producing at least 20 of uninsured funds loss under these assumptions. The aggregate creates a dependency between loss components even though their conceptual identities remain distinguishable.
Recovery from a responsible third party
Suppose that the gross loss associated with a transfer is 400; applicable crime and cyber capacities are 300 and 300. An initial combined indemnity of 350 leaves 50 with the insured. A later third-party recovery of 90, net of recovery costs, must be accounted for under the governing recovery allocation rules. The original payment plus the later recovery is 440, but this figure combines transfers before ultimate adjustments.
Under a purely hypothetical insured-retention-first arrangement, 50 of the later recovery reimburses the insured, leaving 40 to reimburse the insurers; policyholder net recovery equals 400 and insurers' combined net expenditure equals 310. Under a different lawful allocation the insurers may receive more of the recovery before the insured's residual portion is paid. The calculation must reflect the relevant contract and local rules, and the example imposes no universal priority between insurer and insured.
If one insurer initially paid 300 and the other 50, then under a hypothetical equal contribution of 175 each on the initial indemnity, the second would owe 125 to the first. This inter-insurer transfer is separate from the third-party recovery. Recording both as new indemnity paid to the insured would overstate recovery and distort the loss record.
Retentions, collectibility and attachment
Let a first-party loss equal 500. Assume policy A has a retention of 100 and a maximum payment of 250. Under the simplified terms its independent payment is 250 after applying the retention. Policy B, assumed genuinely excess above a specified 350 loss attachment, responds to 150. Total recovery is 400 and the insured bears 100.
Now assume the first insurer is insolvent and cannot pay. Under an assumed excess policy with a fixed loss attachment and no drop-down obligation, the result can be very different from a policy that responds over collectible underlying amounts. If the former otherwise remains obliged for 150, the insured bears 350 rather than 100. A hypothetical policy that explicitly drops down to pay an otherwise covered shortfall could provide greater recovery, subject to its terms. The policy's collectibility, exhaustion and insolvency wording and applicable law determine whether such a scenario is possible.
The use of a retention creates another potential ambiguity. A policy might expressly provide that the other insurer's retention is included within the amount above which it attaches. Another could calculate its own retention in relation to a differently defined loss. Two retentions do not necessarily combine additively or reduce to their larger value. Each must be applied according to the governing wording before the net position is determined.
Defense within and outside limits
Assume a common damages loss of 550 and covered defense expenditure of 150. Two policies have damages limits of 300 and 400. Suppose A's defense is inside its limit and B's defense is outside. If A pays 100 of defense and B pays 50, A has remaining indemnity capacity 200 while B has 400. Damages capacity is then 600, so the 550 damages can be fully paid under an assumed permissible allocation.
If both insurers' defense costs erode their respective limits and they still pay the same 100 and 50 for defense, residual capacities are 200 and 350. Damages exactly exhaust the remaining total. If the damages increase by 50, there is a 50 unpaid damages portion, absent any additional coverage. The comparison illustrates why tracking nominal damages limits without defense erosion can misstate effective coverage.
The exercise also illustrates the need to record which party paid outside counsel and whether those payments were owed under the specific policies. If A incurred expense to contest its own coverage obligation rather than to defend the insured against the underlying claim, that expenditure is not automatically part of the assumed common defense amount.
Empirical identification
Published cases reveal possible conflicts, but their financial amounts and outcomes reflect the selection of disputes that become litigated and accessible. Claims resolved without litigation may be absent, confidential settlements may not disclose allocation, and a single appellate decision may concern a small subset of all payments. A citation count is not a count of incidents, insurance obligations, insureds or claim payments.
Market evidence supplies a different scale of observation. The US Government Accountability Office's (2021) cyber insurance market report highlights based on broker-client data indicate an increase in the proportion electing cyber coverage from 26 percent in 2016 to 47 percent in 2020. The denominator is that broker's client population, not all United States enterprises. The NAIC's (2026) cyber insurance market resource center reports approximately $9.46 billion in United States direct written cyber premium for 2025, including alien surplus lines, an increase of 3.5 percent. These are market volume indicators. Neither measure estimates the incidence of contribution claims.
A valid frequency study would require an exposure denominator of eligible linked incident-policy records. A possible outcome measure would be the proportion of insured incidents with at least two independently established obligations for an identical loss component. A separate rate would measure the proportion of those incidents in which a contribution dispute arose. A further statistic could report the average delay from initial eligible claim to undisputed payment. Each requires event linkage, timing and definitions.
An insurer-level dataset can support analysis of expenditure if it includes gross amounts, contribution received and paid, third-party recovery, defense costs and reserves. A program-level dataset is needed to determine whether multiple insurers cover the same economic loss. Insurer-level and program-level records should be linked without double counting transfers. An observation should preserve the payer, payee, reason, date, amount and corresponding loss identifier.
Cross-country comparisons must distinguish reporting thresholds, statutory rules, premium populations, policy forms, exposure measures and currency. Directly comparing the percentage of litigated cyber claims with the percentage of all claims ceded to another insurer would yield a misleading result even if each percentage were accurately recorded. Converting currencies or standardizing inflation does not repair incompatible definitions.
The selected sources do not supply a linked contribution dataset. The absence of an estimate in this research is an evidentiary boundary. It also motivates an empirical design capable of distinguishing overlap from dispute, dispute from collection and collection from final risk-bearing.
Financial, Legal and Institutional Consequences
The insured's liquidity and protection
An enterprise may experience urgent expense immediately after an incident, while coverage negotiations and inter-insurer communication continue for months or longer. The ability to obtain timely covered funds can matter independently of the amount ultimately payable. A policy that is potentially available but subject to a prolonged dispute may contribute less to short-term operational resilience than an equally valuable undisputed contract.
The value of coordinated insurance includes the reduction of uncovered loss, the provision of usable capacity and the timing of payment. If an insured must finance 1,000 of covered loss for one year at an assumed financing cost of 8 percent, simple annual financing expense is 80. This calculation excludes compounding, taxes, fees and differences in drawdown dates; it is illustrative rather than a market estimate. A contract that permits prompt partial indemnity can reduce that expense even if the final allocation among insurers remains unresolved.
A coverage dispute is therefore not described adequately by its ultimate judgment. A claim history should record provisional expenditure, agreed interim payments, contested sums, expenses incurred to secure payment and eventual net recovery. Different parties may bear costs that do not appear in indemnity alone.
Insurer reserves and recoverables
The gross claim obligation and the ultimate expense after contribution need different records. Assume that an insurer pays 600 and later estimates that another insurer will contribute 250. An anticipated net expense of 350 follows if the amount is collected in full and collection expenditure is ignored. If the eventual contribution is only 150, its net expenditure is instead 450. This is an economic illustration; the accounting recognition, valuation and presentation of such rights depend on the applicable financial reporting and prudential rules.
Contribution valuation depends on legal enforceability, coverage evidence, collection prospects and time. A gross payment does not automatically translate into a certain receivable. A reserve calculation that assumes every overlapping insurance clause will be resolved in favor of contribution risks understating net exposure. Conversely, treating collectable contribution as permanently absent may misrepresent expected net burden. The distinction requires documented assumptions rather than mechanical offsets.
Exposure measurement can also be distorted by organizational structure. The same insurance group may have cyber and commercial crime portfolios; overlapping direct insurance claims could produce internal rather than external economic transfers at the consolidated level. That possibility does not eliminate the need for contractual and accounting attribution at the entity level. Likewise an insurer may transfer part of its cost through reinsurance, which requires a separate analysis of the covered event, underlying allocation and reinsurance contract.
Concentration and hidden overlap
The 2021 New York circular treats systemic cyber risk and non-affirmative exposure as underwriting concerns. EIOPA (2022), in its non-affirmative cyber exposures statement, calls attention to supervisory management across affected products and lines. The insurance-system implication is that an insurer can have exposure to cyber-related economic outcomes even when the relevant exposures are spread across traditional classifications.
One cloud service interruption can affect a large number of firms. Each firm may hold several policies and may have liabilities to third parties also insured by other carriers. Exposure aggregation consequently operates along at least two useful analytical axes: the distribution of costs across insured organizations and the potential duplication of insurance obligations for the same cost within each organization's program. Those are different measurement problems.
A portfolio analysis that counts each potentially responsive policy's full headline limit may overstate realized gross loss for an individual incident, yet understate uncertainty about allocation, timing and defense. A portfolio analysis that counts only affirmative standalone cyber policies may omit possible non-affirmative claims under other classes. The appropriate treatment depends on the intended measure: contractual maximum, scenario loss, expected ultimate net cost, or liquidity requirement.
The cyber-risk literature supplies reasons to examine dependence and information scarcity carefully. It does not supply an empirical coefficient for the effect of other-insurance clauses on aggregate loss. Such a coefficient would require an exposure database and an identified relationship between policy overlap, legal interpretation, loss severity and collection.
Program design and contract procurement
The insurance program is an arrangement of legal obligations as well as a collection of limits. A comparative review of the program can identify which contracts address transferred funds, electronic data, dependent disruption, notification, privacy liability and professional services. It should also consider triggers, responsible insurers, exclusions, retention, attachment, and the treatment of defense.
An enterprise may buy a second contract to increase capacity or broaden coverage. The economic usefulness of that contract depends on what remains covered when the rest of the program is taken into account. A policy described as excess may be expected to attach seamlessly above a specified underlying policy, yet its exhaustion wording might create an unintended gap if the underlying insurer makes a compromise payment or becomes unable to pay. A policy described as primary may contain a broad competing other-insurance provision.
A transparent schedule of available insurance and deliberate layering arrangements can reduce uncertainty about what the parties intended to purchase. Such a schedule remains only one piece of evidence. An endorsement specifying priority in one insurer's contract does not necessarily bind another insurer. For negotiated coordination to have its intended effect, the relevant parties and applicable forms need to support it and any statutory restrictions must be respected.
Incident response and insurance evidence
NIST (2025) integrates incident response into broader cybersecurity risk management in its incident response community profile. This technical framework provides a context for preserving the chronology of events, identifying dependencies and organizing response activities. It does not establish which expenditure should be indemnified under a specific policy.
For insurance research, available incident records have to be connected to financial and contractual records. Logs can indicate unauthorized activity, communications can show when instructions or claims were received, and operational records can quantify stages of interruption. Invoices and payment records identify money spent. The policy and legally relevant notice establish the path from these technical and economic facts to the coverage analysis.
A single consolidated evidence ledger can preserve the relationship between each invoice and the policy or policies asserted to cover it. That record does not presume any final choice of allocation method. It permits the parties to trace disputes through the factual history, independently test calculations and correct entries when amounts or coverage change.
The legal and operational interests can diverge over urgency, preservation and disclosure. Information relevant to coverage may be confidential or privileged, and claim files may contain personal information about affected individuals. A research dataset should respect applicable rules, minimize retained personal data, and separate identifying information from anonymized analytic records where feasible.
Market and policy implications
Clearer coverage language can reduce uncertainty over overlap, but complete elimination of concurrent insurance would sacrifice some of its potential benefits. Multiple available indemnity sources can be economically valuable when individual limits are insufficient, while deliberately arranged layers can achieve higher protection at an appropriate price. The objectives of insured protection and stable pricing are served by clarity of the intended relationship rather than by uniform exclusion of all possible overlap.
Supervisory reporting of affirmative cyber premium is useful but insufficient for understanding inter-line allocation. A more granular report would distinguish first-party and third-party components, standalone and package placements, contribution transfers, recoveries from non-insurer parties, and uncertain non-affirmative claims. The public release of aggregate data should preserve confidentiality and avoid disclosure of identifiable enterprise security vulnerabilities.
Legal diversity is relevant to market design. An insurer operating internationally cannot assume that a priority condition enforceable in one jurisdiction has the same effect elsewhere. The question includes mandatory insurance rules, contract law, governing law, jurisdiction, the identity of contracting parties, and the permitted operation of multiple-insurance provisions. Global policy wording may require local qualification or clarification.
Integrated Scientific Synthesis and Research Agenda
Interpreting the comparative record
The selected sources expose a recurring practical problem: multiple independently purchased obligations may overlap, and allocation rules must operate alongside the insured's actual entitlement. This recurrence does not mean the legal foundations coincide. California's equitable contribution reasoning, the Canadian treatment of competing excess conditions, the English preliminary broker-negligence issue, and the Australian and German statutory schemes emerge from different authorities.
The comparison supports an analytical ordering for research. Establish each obligation on its own terms, determine whether the obligations concern a common compensable loss and comparable levels, inspect contractual coordination and applicable law, and evaluate the effect on the insured's entitlement and on inter-insurer transfers separately. This ordering is a practical consequence of the evidence reviewed. It does not constitute a new doctrine or claim to displace local legal tests.
Three analytically distinguishable questions can therefore be posed for any observed payment. Which policy legally owes money to the insured or on the insured's behalf? Which insurance obligation, if any, is shared with a different insurer? What rights of redistribution arise after expenditure? Resolving any one question does not settle the others. A claim may be payable with no contribution; a payment may later be adjusted; and an external recovery may reallocate the final economic loss without establishing any overlap in original insurance.
The approach avoids treating the specialty insurance label as a proxy for contractual primacy. A cyber insurer can be the only insurer for certain costs, one of several primary insurers for a different cost, or an excess insurer for another. The appropriate outcome may change within a single event. This conclusion follows from observed modular policy structures and the legal treatment of common obligations.
Counterexamples and limits
Counterexample one: two policies name the same cyber event in their conditions but one covers customer damages while the other covers the insured's own restoration expense. Shared event terminology does not make the losses identical. If no common expenditure is insured, a contribution claim cannot be assumed.
Counterexample two: two insuring agreements cover the same damages but one policy is a specifically purchased vertical layer with an unfulfilled attachment. The shared loss does not itself establish a present co-primary obligation. The layering arrangements have independent evidentiary significance.
Counterexample three: both insurers would otherwise cover the same amount but competing other-insurance provisions interact with a mandatory statute. A priority calculation based exclusively on the text, without considering that statute, may be invalid.
Counterexample four: an insurer could have covered the claim had timely notice been given but valid contractual or legal grounds now preclude recovery. A counterfactual claim against a broker or other party for failing to preserve cover may be materially different from a live contribution claim among insurers. The Watford litigation is a reminder of the importance of keeping those relationships distinct.
Counterexample five: an insurer pays a claim and later receives money from a negligent service provider. This recovery reduces its net expense but does not establish that the service provider was a concurrent insurance obligor. Statistical reporting should identify the type of recovery separately.
These counterexamples specify the conditions under which a simplified equal-shares or proportional algorithm becomes inappropriate. They also explain why model inputs must be derived from tested policy obligations instead of assigning every policy a weight based on its headline limits.
Research design for future observations
A longitudinal study could follow anonymized enterprise incident cohorts from first report through claim closure and subsequent recovery. Sampling should include incidents with no insurance claim, incidents with a single responsive policy, incidents with multiple independently responsive policies and actual contribution disputes. Sampling only disputed claims would estimate patterns in litigation rather than incidence across all insured incidents.
For each policy, the data would record coverage section, trigger type, form version, policy period, aggregate, attachment and retention, other-insurance wording, reporting and settlement dates, governing jurisdiction and coverage decision. For each expenditure, the data would record the insured entity, insured interest, loss category, amount, valuation basis and payment relationship. Contribution accounting would link the initial disbursement to subsequent inter-insurer payments. This structure permits the researcher to distinguish incident size from degree of policy overlap.
A central difficulty is confounding. Firms exposed to larger technology risks may buy larger and more numerous policies and also have more severe claims. An association between policy count and disputes could therefore reflect exposure severity, complexity or firm size rather than the effect of overlapping conditions. Comparing outcomes requires the relevant covariates and a defensible sampling design.
Potential dependent variables include time to first payment, outstanding uncompensated covered cost, net insurer expenditure, and incidence of formal priority dispute. Explanatory variables can include policy count, number of relevant coverage grants, use of conflicting excess provisions, severity and jurisdiction. Any statistical study would need to address censoring of still-open claims, missing settlements and the dependence of observations sharing an event or technology provider.
A qualitative complement would collect policy wordings across underwriting years and jurisdictions. The analysis could track changes in conditions, carve-outs for specifically excess insurance, fraud coverage and aggregation provisions. It would test whether changes in wording are followed by changes in observed disputes while controlling for changes in market composition, security exposures and claim severity. Without these controls, an apparent association could result from unrelated market evolution.
The proposed research agenda is motivated by the current evidence boundary. No results of those prospective studies are claimed. The numerical examples use stated capacities; no simulated relationship is presented as an estimated market effect.
Institutional documentation
Insurers and insured enterprises possess different pieces of the information needed for research. An insurer knows its own terms and payments; the enterprise may know the broader program; brokers may possess the purchase history; and legal proceedings may reveal particular contested clauses. A research partnership that links these records under appropriate confidentiality protections would have greater explanatory power than an isolated claims collection.
Consistent terminology is a prerequisite. In particular, a contribution payment, a subrogation recovery, a refund from a vendor, an indemnity installment, and a reinsurance recovery should be identified separately. If a data source labels all these transactions as recoveries, the analyst cannot use their sum as a measure of other-insurance contribution without reclassification.
Jurisdiction-specific documentation must preserve the law applied to each contract and the authority used to resolve a priority issue. The date of the loss and the date of the relevant legal decision are different variables. The study of developments across years requires both. A policy form published later cannot be assumed to govern an earlier claim.
The research results can inform underwriting, portfolio aggregation and scholarly understanding of insurance coordination. Their significance lies in improving the identification of the loss and the relationships among contractual obligations, rather than establishing a prescribed universal insurance arrangement.
Conclusion
Cyber insurance participates in enterprise insurance programs that may contain overlapping, separate, or vertically layered obligations. Determining priority requires identification of the insured, interest, loss component, coverage grant, temporal trigger and applicable law. Other-insurance provisions govern aspects of the relationship between concurrently relevant policies. Contribution concerns the allocation of a shared insurance burden; subrogation and other recoveries arise through distinct legal relationships.
The comparative record demonstrates that statutes and courts can resolve the apparent circularity of competing other-insurance conditions, preserve recovery capacity for an insured, or recognize distinct claims between insurers. It also demonstrates legal diversity. The reported Watford dispute gives a directly cyber-related illustration of simultaneous coverage and broker-notification consequences. Canadian and American cases clarify selected equitable and interpretive mechanisms, while the Australian and German statutes show the importance of specific legislative rules.
The deterministic examples establish that equal shares, proportional weights, sequential attachment and aggregate erosion can produce materially different insurer costs under identical economic loss assumptions. Whether a given computation is lawful depends on the applicable obligations and legal rule. Payment to the insured, redistribution among insurance entities and external recovery must be separately recorded to preserve the economic identity of the claim.
Market statistics and the literature on cyber risk establish the importance of the field and the challenge of limited data. They do not provide the incidence or causal effects of contribution disputes. A linked incident, policy, expenditure and payment dataset is necessary for that purpose. Future comparative evidence should measure those relationships directly and retain jurisdiction, policy edition, exposure population and uncertainty.
The principal substantive result of this study is the separation of loss identification, coverage determination, priority, and ultimate financial allocation within an enterprise cyber insurance program. It explains why an incident-level label cannot settle a loss-component-level issue and why multiple insurance capacity need not imply excess recovery by the insured. These conclusions provide a grounded scientific account of the interactions among cyber, crime, property and liability insurance, within the limits of the sources examined.
References
At-Bay (2023). Cyber Insurance Policy, AB-CYB-001.2, edition 08/2023, pp. 33–34, §§ VII.J–L. https://www.at-bay.com/wp-content/uploads/2023/06/Cyber-Insurance-Policy-Form.pdf
Watford Community Housing Trust v Arthur J. Gallagher Insurance Brokers Ltd [2025] EWHC 743 (Comm), April 8, 2025, paras. 1–12 (accessible extract). https://vlex.co.uk/vid/watford-community-housing-trust-1076458727
zoomLaw (2025). Watford Community Housing Trust v Arthur J. Gallagher Insurance Brokers Limited, [2025] EWHC 743 (Comm), case summary. https://zoomlaw.co.uk/2025EWHC743Comm.html
Simmons & Simmons (2025). Double insurance – the role of ‘other insurance’ clauses. https://www.simmons-simmons.de/en/publications/cmawm1g5800pcvch07rimwtxl/double-insurance-the-role-of-other-insurance-clauses
Fireman’s Fund Insurance Co. v. Maryland Casualty Co., 65 Cal. App. 4th 1279 (Cal. Ct. App. 1998). https://law.justia.com/cases/california/court-of-appeal/4th/65/1279.html
Selective Way Insurance Co. v. Fireman’s Fund Insurance Co., No. 753, September Term 2021 (App. Ct. Md. 2023). https://law.justia.com/cases/maryland/court-of-special-appeals/2023/0753-21.html
Commonwealth of Australia (1984, compilation October 8, 2026). Insurance Contracts Act 1984, §§ 45, 76. https://www.legislation.gov.au/C2004A02944/2026-10-08/2026-10-08/text/original/epub/OEBPS/document_1/document_1.html
Zurich Australian Insurance Ltd v Metals & Minerals Insurance Pte Ltd [2009] HCA 50, December 2, 2009. https://www.hcourt.gov.au/cases-and-judgments/judgments/judgments-1998-current/zurich-australian-insurance-ltd-v-metals-minerals-insurance-pte-ltd
Supreme Court of Canada (2002). Family Insurance Corp. v. Lombard Canada Ltd., 2002 SCC 48, [2002] 2 SCR 695, official reasons and case headnote. https://decisions.scc-csc.ca/scc-csc/scc-csc/en/1988/1/document.do
Federal Republic of Germany (2007). Versicherungsvertragsgesetz (VVG), § 78, official English translation, Insurance Contract Act 2008. https://www.gesetze-im-internet.de/englisch_vvg/englisch_vvg.html
American Tooling Center, Inc. v. Travelers Casualty & Surety Co. of America, 895 F.3d 455 (6th Cir. 2018), July 13, 2018. https://law.justia.com/cases/federal/appellate-courts/ca6/17-2014/17-2014-2018-07-13.html
Merck & Co., Inc. v. ACE American Insurance Co., Nos. A-1879-21 and A-1882-21 (N.J. Super. Ct. App. Div. May 1, 2023). https://law.justia.com/cases/new-jersey/appellate-division-published/2023/a-1879-21.html
New York State Department of Financial Services (2021). Insurance Circular Letter No. 2: Cyber Insurance Risk Framework, February 4, 2021. https://www.dfs.ny.gov/industry_guidance/circular_letters/cl2021_02
European Insurance and Occupational Pensions Authority (2022). Supervisory statement on the management of non-affirmative cyber exposures. https://www.eiopa.europa.eu/publications/supervisory-statement-management-non-affirmative-cyber-exposures_en
United States Government Accountability Office (2021). Cyber Insurance: Insurers and Policyholders Face Challenges in an Evolving Market. GAO-21-477, May 2021. https://www.gao.gov/products/gao-21-477
National Association of Insurance Commissioners (2026). 2026 Report on the Cybersecurity Insurance Market, 2025 reporting year, resource center. https://content.naic.org/resource-center/2026-report-cybersecurity-insurance-market
Biener, C., Eling, M., & Wirfs, J. H. (2015). Insurability of cyber risk: An empirical analysis. The Geneva Papers on Risk and Insurance—Issues and Practice, 40(1), 131–158. https://ideas.repec.org/a/pal/gpprii/v40y2015i1p131-158.html
Eling, M., & Schnell, W. (2016). What do we know about cyber risk and cyber risk insurance? The Journal of Risk Finance, 17(5), 474–491. DOI: 10.1108/JRF-09-2016-0122. https://ideas.repec.org/a/eme/jrfpps/jrf-09-2016-0122.html
National Institute of Standards and Technology (2025). Incident Response Recommendations and Considerations for Cybersecurity Risk Management: A CSF 2.0 Community Profile. NIST SP 800-61 Rev. 3, April 2025. https://csrc.nist.gov/pubs/sp/800/61/r3/final