Guide Cybersecurity Risk Management

Cyber insurance: requirements, exclusions and how to prepare a claim

Buying cyber insurance takes more than choosing a limit and paying the premium. Insurers may assess controls such as multi-factor authentication, endpoint protection, backups and incident response to decide whether to offer cover and on what terms. After an incident, the information given in the proposal, the policy wording and the applicable law can all affect how a claim is assessed. This guide explains how to prepare security evidence, review exclusions and organise the procedures you need if you ever have to notify your insurer.

3 May 2026
Updated 18 November 2026
13 min read
Key takeaways
  • Insurers frequently assess MFA, EDR, backups, privileged access and patching, but requirements vary by insurer, organisation size, sector, geography and limit
  • Answer proposal questions accurately, disclose material qualifications and keep the evidence behind each answer
  • Under the Insurance Act 2015, the consequence of an unfair presentation depends on whether it was deliberate or reckless and on what the insurer would otherwise have done
  • Lloyd's Market Bulletin Y5381 set minimum standards for state-backed cyber attack exclusions; it did not require every state-linked loss to be excluded
  • Insurer notification periods come from your policy, not a universal 72-hour rule, and are separate from reporting to the ICO

Why cyber underwriting changed

Ransomware losses in 2020 and 2021 were higher than many insurers had priced for. Incidents such as Colonial Pipeline and JBS Foods showed how quickly a single attack could produce business interruption, recovery and liability costs. Insurers responded by raising rates, reducing capacity for some risks and asking far more detailed questions about security controls.

The rate increases were steep. In its Global Insurance Market Index for the fourth quarter of 2021, Marsh reported that cyber insurance pricing rose by 130% year on year in the US and by 92% in the UK. Those are averages for two markets, not a global figure, and individual renewals varied. Some insurers also introduced separate sublimits, coinsurance or higher deductibles for ransomware losses.

$769K
Median ransom payment reported in the Sophos State of Ransomware 2026 survey
$1.7M
Average (mean) cost to recover from a ransomware attack, excluding any ransom paid (Sophos 2026)
130%
Year-on-year rise in US cyber insurance pricing in Q4 2021 (UK: 92%), Marsh Global Insurance Market Index

The ransom and recovery figures come from Sophos State of Ransomware 2026, a survey of 2,158 IT and security leaders in 17 countries whose organisations were hit by ransomware in the previous year. Pricing has moderated since the 2021 peak and conditions differ by market and quarter, so ask your broker for current figures. The underwriting discipline that emerged from that period has largely stayed: insurers still expect evidence that key controls are in place, and renewal questions remain detailed.

Five controls frequently assessed in underwriting

Proposal forms can run to dozens of questions. The five controls below are among those insurers most often ask about in detail, particularly in relation to ransomware. They are not universal prerequisites: what an insurer asks for, and how much weight it gives each answer, varies by insurer, organisation size, sector, geography and the limit requested. Depending on the insurer, a gap may lead to a declined quote, different terms, a higher premium or specific exclusions.

1
Multi-factor authentication (MFA)
Insurers commonly ask about MFA separately for email, remote access (VPN and remote desktop), privileged accounts and cloud administration portals. Some insurers decline to quote, or offer restricted terms, where remote access is not protected by MFA. Answer for each system rather than with a single yes or no.
2
Endpoint detection and response (EDR)
EDR tools detect suspicious behaviour and can isolate a compromised device before an attacker moves further. Questions often cover which device types are protected and what share of the devices in scope are covered, so record where EDR is deployed and where it is not.
3
Protected, tested backups
Immutable and offline backups are different things. Immutable storage prevents recovery copies from being changed or deleted for a set retention period; offline or logically isolated copies cannot be reached from the production network. Many organisations combine both. Whatever the design, test restores regularly, record the results and protect backup administration credentials, because attackers often target the backup console.
4
Privileged access management (PAM)
Controls over who holds administrator credentials, such as separate admin accounts, credential vaulting and time-limited elevated access. Privileged accounts are a common route for ransomware operators to move across a network, which is why insurers ask how they are managed.
5
Vulnerability and patch management
Insurers often ask for your target timeframes for applying critical and high-severity patches, and whether internet-facing systems are treated separately. Questions and expected timeframes differ between insurers. If a known unpatched vulnerability was the route in, that may be relevant to a claim, but the effect depends on what you declared, the policy wording and the applicable law.

Beyond these five, insurers increasingly ask about email authentication (SPF, DKIM and DMARC), network segmentation, security awareness training, and whether you have an incident response plan that has been tested recently.

Answering the proposal form accurately

Accuracy matters as much as the controls themselves. If an incident later reveals a gap between what was in place and what the proposal stated, the insurer may raise it when assessing the claim. Whether that affects cover depends on what was represented, how material it was, the policy conditions and exclusions, what you knew, any connection between the gap and the loss, and the applicable law.

Typical application questions cover:

  • Whether MFA is deployed on email, VPN, remote access, cloud services and privileged accounts, asked separately rather than as a single yes or no
  • The percentage of endpoints covered by EDR
  • Your patching timeframes for critical vulnerabilities
  • Whether you have a documented incident response plan and when it was last tested
  • How backups are protected, and whether and how often restores are tested
  • Whether you have PAM controls, and whether privileged credentials are vaulted
  • Annual revenue, number of employees, data types processed (personal data, payment card, health) and sector

Three habits reduce the risk of a dispute later:

  • Answer accurately. Do not round up. If MFA covers most but not all remote access, say so and explain the exception.
  • Disclose material qualifications. Exceptions, legacy systems, pending projects and compensating controls belong in the answer or in a covering note to the insurer.
  • Ask about ambiguous questions. If a question could be read more than one way, ask your broker or the insurer how it should be interpreted and keep a record of the reply.

Do not complete the questionnaire from memory. Build an evidence register before you start: for each question, record the exact wording, your answer, its scope or qualifications, the supporting evidence, who reviewed it and the date it was approved.

Illustrative example. The register below shows the format only. The entries are fictional and do not reproduce any particular insurer's questions.

QuestionAnswerScope or qualificationEvidence
Is MFA enforced for all remote access?Yes, with one exceptionEnforced on VPN and remote desktop gateway; one legacy supplier connection is restricted by IP allow-list until its planned migrationConditional access policy export, VPN configuration, migration plan
Is EDR deployed on all endpoints?98% of in-scope devicesLaptops and servers covered; a small set of lab devices without EDR are on an isolated networkEDR console coverage report, asset inventory
Are backups protected against deletion or encryption?YesDaily backups to immutable storage with fixed retention, plus a monthly offline copyStorage configuration, latest restore test report
Within how many days are critical patches applied?14 days internet-facing, 30 days internalExceptions approved through change managementPatch compliance report, exception register
Practical tip

Request a copy of last year's proposal from your broker. Comparing year-on-year answers against your current posture is one of the fastest ways to identify where your documentation does not match your controls.

For business insurance in the UK, the main disclosure rules come from the Insurance Act 2015. The Consumer Insurance (Disclosure and Representations) Act 2012 applies to consumers, not to businesses buying cyber cover.

Before the contract is entered into, the insured must make a fair presentation of the risk (section 3). In broad terms, that means disclosing every material circumstance it knows or ought to know, or at least giving the insurer enough information to prompt further enquiries, in a reasonably clear and accessible way. Material representations of fact must be substantially correct, and representations of expectation or belief must be made in good faith.

A breach does not automatically cancel cover. Under Schedule 1, the insurer has a remedy only if it shows that, with a fair presentation, it would not have entered into the contract or would have done so on different terms. The remedy then depends on the type of breach:

  • Deliberate or reckless. The insurer may avoid the contract, refuse all claims and keep the premium.
  • Neither deliberate nor reckless. The remedy reflects what the insurer would have done: if it would not have written the policy, it may avoid it but must return the premium; if it would have imposed different terms, the contract is treated as including them; if it would have charged a higher premium, the claim may be reduced proportionately.

Section 11 also limits reliance on certain terms intended to reduce the risk of a particular kind of loss, where the insured shows that the non-compliance could not have increased the risk of the loss that actually occurred. Business policies can vary some of these default rules if the Act's transparency requirements are met, so read the wording and take legal advice on any specific dispute.

How insurers assess claims

When you notify a claim, the insurer needs to establish what happened and whether it is covered. That can involve document review, technical investigation, loss adjusting and checking the policy conditions. Many policies give access to a panel of approved incident response, forensic and legal specialists, and some require or encourage you to use them.

Forensic investigators establish the cause, scope and impact of the incident: how the attacker got in, which systems and data were affected, and what recovery will take. Their findings may also inform the coverage assessment, for example where the route in relates to a control described in the proposal. That is one reason accurate answers and good records matter.

Notification is the second issue. There is no universal deadline: each policy sets its own triggers and periods, such as notice as soon as practicable after you become aware of an incident or circumstance, sometimes with a stated number of days. Some notification clauses are conditions of cover. Record the insurer's contacts, triggers and periods in your incident response plan so the team can follow them under pressure.

Notifying your insurer is separate from regulatory reporting. Under UK GDPR, a personal data breach that is likely to pose a risk to individuals must be reported to the ICO without undue delay and, where feasible, within 72 hours of becoming aware of it. Sector regulators and contracts may impose their own deadlines. One notification does not cover the others.

Important

Where feasible, check your policy and contact the insurer or its breach response line before appointing PR, legal or forensic providers, because some policies only reimburse approved providers or require prior consent for costs. Those requirements must not delay mandatory legal notifications, urgent steps to protect people and systems, or containment. Record what you did and why.

Exclusions and cover limits to understand

Read the full policy wording and schedule, not just the summary. The terms that matter most in a claim often sit in the definitions, exclusions and conditions.

!
War and state-backed cyber attack exclusions
Lloyd's Market Bulletin Y5381, issued in August 2022, required standalone cyber policies written at Lloyd's from 31 March 2023 to exclude war losses (unless a separate war exclusion applies) and losses from state-backed cyber attacks that significantly impair a state's ability to function or its security capabilities. Clauses must also be clear about systems outside an affected state, set out a robust basis for attributing an attack to a state and define key terms. The bulletin did not require exclusion of every loss linked to a state-backed actor. Model clauses such as the LMA 5564 to 5567 family differ in scope and in how attribution works, so check your own wording.
!
Ransomware sublimits
Some policies cap ransomware-related losses at a sublimit below the overall limit, or apply coinsurance. A policy with a £5 million headline limit might, for example, cap ransomware losses at £1 million. Check the schedule for every ransomware and extortion sublimit.
!
Prior known incidents and circumstances
Policies commonly exclude, or deal separately with, incidents and circumstances known before the policy started. Disclose known incidents as the proposal asks. Whether a related loss is covered depends on the wording, what was disclosed and when you became aware of it.
!
Maintaining declared controls
Some policies include conditions about maintaining security controls during the policy period. If a declared control was removed or weakened, the insurer may raise it on a claim; the effect depends on the wording, whether the change was disclosed, its connection to the loss and the applicable law. Tell your broker or insurer about material changes.
!
First-party and third-party cover
What is covered depends on the insuring agreements, exclusions and sublimits. Some policies cover regulatory defence costs. Whether fines, penalties and ransom payments can be covered depends on the policy terms and on whether they are insurable under the applicable law.

Cyber insurance readiness checklist

Use this checklist to keep the evidence behind your proposal answers in one place, and update it whenever controls change rather than only at renewal.

AreaWhat to document
PolicyCurrent wording and schedule, limits, sublimits, deductibles, exclusions, endorsements and panel providers
MFASystems and user groups covered, enforcement method, exceptions and compensating controls
Endpoint securityEDR deployment by device type, coverage percentage, excluded devices and why
BackupsDesign (immutable, offline, isolated or a combination), retention, latest restore test, protection of backup admin credentials
Privileged accessInventory of admin accounts, vaulting, time-limited access, review dates
Vulnerability managementPatching timeframes by severity, latest scan results, exception register
Incident responseCurrent plan, insurer contacts and notification triggers, regulatory deadlines, latest exercise and findings
Underwriting representationsExact question, answer given, evidence, reviewer and approval date
Claims procedureHow to notify, information the insurer needs, approved providers, consent requirements for costs
RenewalPlanning timeline, broker contact, changes since the last proposal, open remediation items

Preparing for renewal

Starting around six months before renewal is a sensible planning window, not a fixed rule; organisations with complex estates or open remediation work may need longer. Brokers often send pre-renewal questionnaires a few months out. The bigger point is to keep security evidence current all year, so renewal becomes a review rather than a scramble.

1
Update your evidence register
Document what you have, where it is deployed, who is responsible and when it was last reviewed. Do this for every control the proposal asks about, not just the five above.
2
Test your backups
Run a restoration test and produce a written report with date, scope and outcome. A backup log showing scheduled runs is not the same as a tested recovery. Confirm that recovery copies are protected against change or deletion.
3
Audit your MFA coverage
Check each remote access system, cloud console and privileged account interface and verify that MFA is enforced, not merely available. Note any gaps and how they are mitigated.
4
Update your incident response plan
Include named contacts for your insurer or its breach response line, legal counsel and any approved providers. Record the notification triggers and periods from your policy alongside regulatory and contractual deadlines. Document the last review date.
5
Conduct a tabletop exercise
Simulate a ransomware incident with your core team, including the insurer notification step. Document the date, participants and findings. Insurers increasingly ask whether this has been done in the past 12 months.
6
Run a vulnerability scan before applying
Export a current scan showing your patch status against known vulnerabilities. It supports your answers on patching timeframes and reveals critical exposures to close before you submit the proposal.

An organisation that arrives at renewal with a current evidence register, a tested backup report and a documented tabletop exercise is in a stronger position than one relying on memory. The same records provide consistent documentation that can be checked if questions arise after an incident.

ENISA's guidance on minimum security baseline measures is available at enisa.europa.eu. The NCSC publishes practical implementation guidance on the technical controls that underwriters and regulators most commonly ask about.

If you are building a controls inventory for the first time and need a programme structure to sequence the work, a cybersecurity roadmap maps each control to a business priority and assigns ownership before the insurer questionnaire arrives.

Frequently asked questions

Does cyber insurance require MFA?

Many insurers ask detailed questions about MFA, particularly for email, remote access, privileged accounts and cloud administration, and some will not quote, or will restrict cover, without MFA on remote access. It is not a universal requirement, and expectations vary by insurer, organisation size, sector and limit. Answer the MFA questions accurately for each system and keep evidence of enforcement.

Can an insurer refuse a claim because MFA was missing?

It can be relevant, especially if the proposal stated that MFA was in place. It does not automatically mean a claim will be refused or reduced. The outcome depends on what was represented, whether it was material, the policy conditions and exclusions, what you knew, any connection to the loss and the applicable law. In the UK, the Insurance Act 2015 provides proportionate remedies where an unfair presentation was neither deliberate nor reckless. Disputes can go through the insurer's complaints process, the Financial Ombudsman Service for eligible smaller businesses, or the courts.

How quickly must I notify my insurer of a cyber incident?

There is no universal 72-hour rule for insurers. Follow the notification triggers and periods in your policy, which may require notice as soon as practicable or within a stated number of days, and check whether notice is a condition of cover. Record them in your incident response plan. Insurer notification is separate from reporting a personal data breach to the ICO, which UK GDPR requires without undue delay and, where feasible, within 72 hours.

Does cyber insurance cover ransomware?

Many cyber policies cover some ransomware-related costs, such as incident response, forensic investigation, data restoration and business interruption, often subject to sublimits, deductibles or coinsurance. Cover for extortion payments depends on the policy wording, exclusions and whether a payment is lawful and insurable in the circumstances, for example where sanctions apply. Check the ransomware and extortion sections of your schedule before you need them.

What did Lloyd's require for state-backed cyber attack exclusions?

Lloyd's Market Bulletin Y5381, issued in August 2022, set minimum requirements for war and state-backed cyber attack exclusions in standalone cyber policies written at Lloyd's from 31 March 2023. Clauses must exclude losses from state-backed cyber attacks that significantly impair a state's ability to function or its security capabilities, be clear about systems outside an affected state, set out a robust basis for attribution and define key terms. The bulletin did not require exclusion of every loss linked to a state-backed actor, and it does not put the burden on the insured to prove an attack was not state-sponsored. Clause structures, including the LMA 5564 to 5567 family, differ, so check your own wording and its attribution provisions.

How do I document my security controls for a cyber insurance application?

Build an evidence register that records, for each proposal question, the exact wording, your answer, its scope or qualifications, the supporting evidence, the reviewer and the approval date. For MFA, verify enforcement on each system rather than just availability. For backups, run a test restoration and produce a written report. For patching, export a current vulnerability scan. For incident response, note when the plan was last reviewed and whether a tabletop exercise has been held in the past 12 months.

What is the difference between first-party and third-party cyber cover?

First-party sections typically address your own costs, such as incident response, forensic investigation, data restoration, business interruption and notification costs. Third-party sections address claims made against you by others, such as customers and partners. Some policies also cover regulatory defence costs. What is actually covered depends on the insuring agreements, exclusions and sublimits, and whether fines, penalties or ransom payments can be covered depends on the policy terms and on insurability under the applicable law.

Ryland Deakin
About the author
Lead Consultant, Cyvra · CISM · CompTIA Security+ · MCP

Ryland has delivered cybersecurity, compliance, and IT management programmes for regulated organisations across the UK and the Netherlands for over 20 years. View full profile

Talk to Cyvra

Is your security posture ready for cyber insurance renewal?

We help organisations in the Netherlands and UK close the controls gaps that affect insurability and premium costs, with consistent documentation that can be checked after an incident.

Disclaimer: This article is for general informational purposes only and does not constitute legal, regulatory, or professional advice. Cyvra makes no warranty as to the accuracy or completeness of this content, which may not reflect the most current regulatory developments. Readers should seek independent legal and regulatory advice appropriate to their specific circumstances. Cyvra accepts no liability for any loss arising from reliance on this content.